iron ore primer

European Metals & Mining: A Strange Love — How I Learned to Stop Worrying and Love the Ore

JULY 2013

SEE DISCLOSURE APPENDIX OF THIS REPORT FOR IMPORTANT DISCLOSURES AND ANALYST CERTIFICATIONS

Is $80/t iron ore a certainty? Perhaps not...the future may not be as dark as consensus believes

Recent weakness in mining equities is consistent with iron ore at $80/t Aus FOB, or ~40% below both average five-year and 1H:13 prices; we show how the magnitude and speed of the implied price decline express exaggerated fears around GDP, demand growth and supply response

China, the world's largest commodities consumer (56% of 2012 iron ore), must complete its transformation from a U.S. 1930s' equivalent level of industrialization before it faces an unprecedented demographic crunch; this supports iron ore demand for 7-12 years

Maintaining "super-cycle" pricing, thus, depends on the supply side exercising sufficient capex restraint not to displace the bulk of high-cost producers; however, given our view on demand, we see room for the currently approved pipeline and prices supported above $120/t for seven years

What if we are wrong? Even then, if iron ore fell to $80/t, three "automatic stabilizers" — the natural hedge between revenue and costs, value-in-use and iron ore demand elasticity — would offset some of the fall with a positive value impact of 5-20% for our coverage stocks

For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 16-Jul-2013

EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING

AND LOVE THE ORE 1

Portfolio Manager's Summary

Iron ore has been the outperforming commodity of the last decade (rising 1,340% from its 2001 average to its 2011 peak). Since 2009, the iron ore price has averaged close to $130/t FOB Australia; however, recent weakness in mining equities suggests that the market believes that an imminent, severe and permanent downward revision to the value of this commodity is due. An analysis of equity values shows that the market is discounting a 40% correction in the price of iron ore down to levels close to $80/t. We do not believe that such a decline is necessarily justified and that it is, at the very least, inconsistent with the growth forecasts for the Chinese economy held by the IMF and the OECD. We believe that the price of iron ore is supported by the fundamental geological poverty of China and that any reacceleration of the Chinese economy must imply a reacceleration in metal demand and with it support for the price of raw materials.

That China is the world's largest consumer of commodities (56% of iron ore in 2012) is undeniable. Equally undeniable is the unprecedented rapidity with which China has attained this position. However, we also believe that the demographic crunch looming in 2020-25, as the birth-rate explosion of the 1950s and 1960s and subsequent "one child" policy successively age their way through the population, is a phenomenon that cannot be ignored. If China is to industrialize sufficiently to support its aging population before exhausting its supply of "cheap" labor, then it has a bare decade in which to complete a transformation from an agrarian society to one dominated by the service sector. China's embedded capital stock (~4.5 tons of steel per capita) and the composition of its labor force (with parity between primary and tertiary forms of employment) resemble that of the U.S. of the 1930s. Furthermore, while China's rate of capital stock formation is roughly 2x that of the U.S., this merely speaks of the increases in mining productivity that we have seen over the last 50 years. Productivity gains mean that China's capital stock is embedded at one-fourth the labor used to build up that of the U.S.

Maintaining so-called "super-cycle" pricing, thus, depends upon the supply side. Specifically, will the major Western miners exhibit sufficient capital discipline to limit the quantity of low-cost iron ore brought to market? This commodity can be mined by the likes of the "Big Three," for as low as $25-30/t cash costs. Will they race to compete on low-cost volume, thereby displacing the ~9% of 2012 global supply provided by high-cost ($120-$180/t) Chinese miners? Increasingly, the miners are becoming aware that prioritizing volume over value is a fool's paradigm and that just because you can build something does not mean that you should. In this Blackbook, we show how iron ore prices could well be supported above $120/t through to 2020.

Furthermore, even if we are wrong and iron ore prices do decline as the market is anticipating, three factors will offset their impact on company valuations: First, the natural hedge between costs and revenue (including but not limited to exchange rates); second, because not all iron ores are created equal, value-in-use adjustment reduces the ability for low-quality supply to compete and provides price support for the major miners with high-quality deposits; and third, the elasticity of Chinese iron ore demand. We believe that a major part of the slowdown in Chinese steel growth is the budgetary constraint of high iron ore prices; relax the constraint through lower iron ore price and iron ore demand should increase.

Paul Gait [email protected] +44-207-170-0599Esther L. Healer, CFA [email protected] +44-207-170-0546Rusne Didziulyte [email protected] +44-207-170-0541

July 15, 2013


2 EUROPEAN METALS & MINING: A STRANGE LOVE — HOW I LEARNED TO STOP WORRYING

AND LOVE THE ORE



AND LOVE THE ORE 3

Table of Contents

Significant Research Conclusions 5

If China Is to Grow Rich Before It Grows Old, It Must Finish Industrializing in the Next Decade 15

Iron Ore "Super-Cycle" Pricing Generated by High-Cost Producers Responding to the Rapidity of China's Industrialization 29

Two Times the Rate at One-Fourth the Labor: Chinese Steel Intensity Does Not Indicate Overinvestment 41

Iron Ore Pricing Support Over the Next Decade Depends Upon Capital Discipline (Rio 360 Case Study) 63

What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities? 81

China Growth Scenarios 105

Our Iron Ore Price Forecast 115

Local Currency Costs and USD Revenues: The Stabilizing Effect of the "Natural Hedge" in Costs 135

Not All Iron Ores Were Created Equal: The Stabilizing Effect of Value in Use 143

If Prices Fall While Capital Stock Is Accumulating, Demand Accelerates: The Demand Elasticity Stabilizer 153

Valuation and Risks 167

Company Impact 187

Appendix 1: From Grizzly to De Niro 205

Appendix 2: Growth Scenarios 253

Appendix 3: Financial Statements 259

Index of Exhibits 277



AND LOVE THE ORE

Exhibit 1 Financial Overview

Source: Bloomberg L.P., FactSet and Bernstein estimates and analysis.

7/12/2013 Anglo American BHP Billiton Glencore Xstrata Rio Tinto Vale MSCI Europe

AAL.LN BLT.LN GLEN.LN RIO.LN VALE3.BZ MSDLE 15

Rating O O O O O ‐

Local Currency Units £ £ £ £ BRL

Current Share Price (Local Currency) 12.95 18.00 2.60 27.99 30.32 367.33

52 Week High (Local Currency) 21.63 22.36 3.98 37.57 44.10 379.59

Current Low (Local Currency) 12.07 16.71 2.69 25.82 28.39 303.28

YTD Performance (%) (29.5%) 1.3% 7.5% (16.1%) (22.0%) 16.1%

YTD Relative Performance (%) (45.6%) (14.8%) (8.6%) (32.2%) (38.1%) ‐

12 Month Price Target (Local Currency) 20.25 22.50 5.25 41.25 46.50 ‐

Upside/(Downside) to Target Price 56% 25% 102% 47% 53% ‐

MCAP (US$m) 25,217 145,992 27,483 78,780 77,034 ‐

Net Debt/(Cash) (US$m) 5,855 32,169 22,551 21,990 26,088 ‐

Minorities 17,642 1,349 5,032 490 (4,702) ‐

EV (US$m) 51,687 170,890 93,427 98,682 98,420 ‐

EBITDA ‐ US$m

2013E 10,052 25,614 15,902 24,183 22,523

2014E 11,833 38,119 20,203 31,348 26,230

2015E 15,590 45,903 26,252 38,698 32,415

EV/EBITDA ‐ US$m

2013E 5.1 6.7 5.8 4.1 4.4

2014E 4.4 4.5 4.2 3.1 3.8

2015E 3.3 3.7 3.3 2.6 3.0

EPS ‐ US$/share

2013E 1.90 2.70 0.43 5.33 2.28

2014E 2.57 3.73 0.63 7.34 2.55

2015E 3.75 4.73 0.97 9.52 3.58

Dividend Yield ‐ %

2013E 3.5% 4.2% 3.5% 3.9% 2.2%

2014E 2.3% 3.5% 4.9% 4.1% 3.4%

2015E 3.0% 3.6% 5.8% 5.1% 4.8%



AND LOVE THE ORE 5

Significant Research Conclusions

Iron ore has been the outperforming mining commodity for the last decade and is the critical source of earnings generation (~60% of EBITDA) for our coverage group. Recent weakness in mining company equities implies a very severe discount to spot iron ore price ($80/t versus an 1H:13 average US$129/t Australian FOB) in terms of future price level and the rapidity of decline. We do not believe this decline is a foregone conclusion. Furthermore, if it were to happen, there are a number of factors that would mitigate the severity of the valuation impact.

The industrial behemoth1 that is the Chinese economy has been powered by exploitation of China's most abundant resource: vast pools of surplus rural labor. However, this labor supply is not infinite. Over the next generation, China will experience a substantial increase in its old-age dependency ratio as the country ages sharply in response to earlier policy initiatives (namely, the "one child" policy) — see Exhibit 2. On top of this are the behavioral changes in a workforce increasingly composed of only children. The IMF sees cheap labor in China exhausted between 2020 and 2025 — concurrent with our forecast for peak raw materials demand in China (see Exhibit 3). We believe that the proximity of this labor exhaustion has contributed to the urgency of the Chinese governmental push to install a productive capital base and transition the economy away from subsistence agriculture to tertiary forms of value-added activity. The unprecedented rapidity of Chinese industrialization has rendered China the largest consumer of global commodities (56% of total iron ore demanded globally in 2012). In our view, this reflects the country's requirement to "grow rich before it grows old" and before the supply of cheap labor that has powered this industrialization is exhausted.

Exhibit 2 In 2000, China Was a Clear Outlier — the World's Largest Country Set to Age Rapidly and Yet Still at a Pre-Industrial Output Level

Exhibit 3 The Demographic Challenges Facing China Stand Behind the Requirement to Accelerate the Capital Stock Formation

Source: UN and Bernstein estimates and analysis. Note: 2015-40 numbers are UN estimates.

Source: UN and Bernstein analysis.

1 The original biblical Behemoth is described as having a tail like cedar, bones like iron, limbs as strong as copper, and sinews like stone, hence it seems an apropos metaphor for the economy of the world's largest consumer of commodities.

-10

0

10

20

30

40

50

60

-50 -40 -30 -20 -10 0 10

GD

P/C

apit

a —

US

$'00

0

Change in Old-Age Dependency Ratio 2000-2040 — Negative Indicates Aging Population

Aging of Population vs. Development of Economy Including ChinaSize of Bubble = Population

USA

Japan

ChinaSouth Korea

-10%

-5%

0%

5%

10%

15%

20%

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

Ch

ang

e in

Wo

rkin

g-A

ge

Po

pul

atio

n

China — Growth in Working-Age Population (15-64 Years)

Critical demographic inflection point ~2020‐2025

We Do Not Believe the Sustained Iron Ore Price Fall Implied by Mining Equity Valuations Is a Foregone Conclusion; Even If So, Several Factors Would Mitigate the Valuation Impact

If the World's Largest Consumer of Commodities Is to Grow Rich Before It Grows Old, China Must Largely Complete Its Unprecedentedly Rapid Industrialization by 2020-25



AND LOVE THE ORE

It is the urgency of China's industrialization, in our view — coupled with the fundamental geological scarcity of certain key raw materials domestically — that has been the key driver behind the emergence of the commodity "super-cycle." On the demand side, our secular commodity price view is supported by a fundamental analysis of industrialization patterns in 120 countries since 1900. On the supply side, we identify the emergence of a new class of marginal producers in 2003 (Chinese high-cost producers), their impact on the global cost curve, and the duration of supply side support for "super-cycle" pricing based on new greenfield and brownfield projects. We see the "super-cycle" pricing regime supported on the demand side until China finishes industrializing (2020-25) and on the supply side until sufficient new low-cost Western production is brought on line to displace the marginal high-cost Chinese producers (the timing of which rests in the hands of the Western majors and underlies our continued — and continuous — calls for capital discipline) — see Exhibit 4.

The high rate of recent Chinese steel consumption, in both absolute magnitude and relative to the current consumption by other countries, is often interpreted to suggest overinvestment and that the Chinese economy may be unduly steel intensive — in which case, an immediate and sustained iron ore price decline to US$80/t might indeed be reasonable. However, many datasets look at only the last ~20 years and point out China's peak steel intensity of 60kg/000$GDP as being substantially higher than the peak rate of consumption during the industrialization of the U.S. They fail to consider the pronounced productivity increases in the mining sector over the last half-century. China has access to iron ore production that is 2.5-5x more efficient in man-hour terms than was available to the U.S. when it industrialized (see Exhibit 5). Three factors stand behind China’s access to this higher productivity iron ore mining (largely located in Australia): 1) globalization of the commodities industry (between 1965 and 2005, real freight costs fell 40%, while global trade in iron ore rose by 330% and coal by 1070%), 2) geological endowment (a miner in the U.S. needs to move ~2x ROM material as his Australian counterpart to generate the same amount of useful material), and 3) population densities (Western Australia ex-Perth has, on average, 0.23 people per square kilometer versus 67 in Minnesota and 175 in Michigan). Due to this increased productivity and China's access to it, the Middle Kingdom2 is embedding steel into its capital stock at a rate ~2x that of peak U.S. steel intensity — but at one-fourth of the labor that the U.S. required. This more rapid embedding of steel in China's capital stock reflects not inefficiency per se, but rather that the productivity of the world's best iron ore miners (Australians today) can be utilized to hasten the process of capital stock formation. This rapidity has been further enabled by the purchase, disassembly and reassembly of steel-making production from the West into China (e.g., the Famous Industrials Group industry relocation projects). Consequently, from a demand perspective, we consider the fear that the iron ore price will decline to US$80/t and remain there to be overdone.

2 The literal translation of Zhong-guo — the Chinese name for China, which can also be rendered as Middle State. American is Meiguo or Beautiful Kingdom/State.

We Attribute "Super-Cycle" Pricing in Iron Ore to the Emergence of High-Cost Chinese Marginal Producers; We See the "Super-Cycle" Continuing Until Those Producers Are Displaced by New Low-Cost Production

We Do Not Believe That the Rate of Chinese Steel Consumption Is Indicative of Overinvestment; China Is Embedding Steel Into Its Capital Stock at a Rate Approximately Double That of Peak U.S. Steel Intensity But at Around One-Fourth of the Labor That the U.S. Required



AND LOVE THE ORE 7

Exhibit 4 China Accounted for ~16% of 2012 Global Iron Ore Production (ROE Equivalent); Our Analysis Shows That ~60% of the Chinese Production Consisted Primarily of "Missing" Chinese High-Cost Mines

Exhibit 5 Productivity Uplift of the Iron Ore Imports Into China Has Enabled It to Overtake the U.S. in Terms of Effective Productivity in 2003

Note: There are a few other high-cost producers in the "missing Chinese mines" category but they are de minimis.

Source: AME and Bernstein estimates and analysis.

Source: WSA, NBS, BLS and Bernstein estimates and analysis.

On the supply side, we see the inability of the Chinese mining sector to transition from labor-intensive to capital-intensive modes of production during this rapid industrialization phase in China, supporting the current "super cycle" pricing. Costs on the supply side provide the long-term pricing structure, meaning pricing support will persist, in our view, so long as a significant proportion of high-cost marginal producers in China (~9% of 2011 ROE equivalent global supply, with marginal costs of production ~$120-180/t) are not displaced (see Exhibit 4). Accordingly, capital discipline, the single greatest source of value creation in the mining space, is particularly critical for iron ore (the fourth most crustally abundant element).

Here, we provide a case study on the possible 70Mtpa Pilbara iron ore expansion proposed by Rio Tinto. The price elasticity backed out from Rio Tinto's own cost curve implies a reduction of US$27/t from this project, resulting in an $18 billion value loss (not quite Alcan levels, but getting there). However, if management's 3% steel demand growth target came to pass, this would limit price declines from the project to US$14/t — meaning that expenditure of US$5 billion would generate a positive value of ~US$8 billion. Shareholders in Rio Tinto may be comfortable risking an $18 billion loss for the chance of an $8 billion gain, should China achieve 3% steel demand growth in a decelerating environment. Alternatively, they might wish to follow the example of Emperor Augustus and consider the whole expansion akin to fishing with a golden fish hook: the gain, if all goes well, is too small to offset the risk of loss if it does not. We further expand our analysis to demonstrate how globally superior value generation can result from the pursuit of locally optimal growth targets — but only in consolidated markets where industry incumbents are explicitly aware of the price destructive effects of excessive volume growth (see Exhibit 6 and Exhibit 7). It is this dynamic, which underlies our continued calls for capital discipline from the Western majors to avoid engaging in a massive value transfer from their own shareholders to the Chinese Communist Party by subsidizing China's continued industrialization.

170

160

150

140

130

120

110

100

80

70

0

60

180

40

30

20

10

0

50

1,000,000750,000500,000250,000

90

1,250,000

Cumulative global production– kt Fe

Missing Chinese

Mines

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

To

ns

pe

r Man

Ye

ar

Effective Chinese Productivity, Including Imports

USA Effective Chinese Productivity

Pricing Support in the Next Decade Thus Depends Upon Capital Discipline from the Western Miners as Our "Rio 360" Case Study Shows



AND LOVE THE ORE

Exhibit 6 The Value Destructive Effect of Excessive Volume Growth Can Be Seen Once the Impact of Supply Elasticity Is Considered

Exhibit 7 It Is the Elasticity of Supply That Inflects Any Commodity Cost Curve; the Higher the Elasticity, the Higher the Margin and the Lower the Degree of Consolidation Required to Generate Superior Returns

Source: Bernstein estimates and analysis. Source: Bernstein estimates and analysis.

The iron ore price (the most important source of EBITDA generation for our coverage) has averaged US$135/t over the last five years and US$129/t 1H:13 Australian FOB — and yet mining company stocks have been among the worst performers of the year. We ran four fixed commodity price scenarios ( "Grizzly," "Bear," "Bull," and "De Niro" in order of increasing aggressiveness) and examined how sensitive the value generated by our DCF (out to 2030) was to these different scenarios. We then selected the scenario that most closely resembled current market conditions and flexed the prices of the two most important commodities (from an EBITDA generation perspective) — copper and iron ore — in order to determine what the market was pricing in for those commodities. We found that mining equity valuations were baking in $80/t iron ore, a ~40% discount to year-to-date and five-year average prices. We then asked ourselves what the world needs to look like for these prices to come about. Specifically, "what growth outlook for the world's largest commodity consumer, China, would be consistent with $80/t iron ore?" We calculated the Chinese GDP growth implied by consensus demand expectations in China, then used this implied economic growth to back out the supply side response that, in concert with this demand expectation, would result in $80/t iron ore. We found that this was consistent with ~7% Chinese GDP growth — some 1.3% below the IMF forecast and 0.5% below the official government guidance — as well as significant new supply: ~550Mtpa of new iron ore capacity out to 2020 (a ~50% increase on the 1,123mt produced in 2012 globally), of which ~300Mtpa would be surplus at this GDP growth rate and hence price destructive.

Using the model (discussed earlier) that we built to back out Chinese GDP and supply side response consistent with consensus demand and price expectations, we can also determine a consistent forecast for iron ore prices and steel demand growth across any GDP scenario for China. Our "bull" scenario takes the IMF's forecast from 2Q:12, which sees the Chinese economy accelerate back up to a trend rate of 8.5%. This would absorb the ~300Mtpa of price destructive oversupply implied in the consensus scenario (mentioned earlier), resulting in a tight market in which the miners do not compete on volume and in which a trend price of US$140/t could

0

20

40

60

80

100

120

140

0 25 50 75 100125150175200225250275300

Val

ue

(US

$bn

)

Iron Ore Volume Growth for New Entrant (Mt)

Value of Volume Growth for Different Price Elasticity

0.0 -0.5 -1.0 -1.5-2.0 -2.5 -3.0

"Normal" straight‐line relationship between volume and value...butonly if elasticity is

ignored

0

2

4

6

8

10

12

1 2 3 4 5 6 7 8 9 10EV/E

BIT

DA

Mu

ltip

le o

f In

du

stry

No. of Players in Industry (i.e., Degree of Consolidation)

Multiple Uplift Available in Consolidated Commodity Markets by Elasticity

e=-1 e=-2 e=-3

Mining Equities Have Recently Been Baking in $80/t Iron Ore (~40% Below Year-to-Date and Five-Year Average Prices) — This Is Consistent With Both ~7% Chinese GDP Growth (Below Both IMF Estimates and Chinese Targets) and ~550Mtpa New Iron Ore Capacity Out to 2020

We Construct Bull, Middle and Bear Scenarios for Chinese GDP and Supply Response to Bound the Possibilities for Iron Ore Prices Out to 2020 ($140/t down to $65/t)



AND LOVE THE ORE 9

eventuate out to 2020 (see Exhibit 8). A "middle" scenario takes the Chinese target of 7.5% GDP growth as a floor (despite the fact that Chinese GDP has a tendency to miss on the upside). This yields a price of ~US$120/t out to 2020. Our China "bear" scenario sees GDP decline to 6% and no supply coming out of the market (a true worst-case scenario). This results in an iron ore price testing the US$65/t mark over a multi-year period.

Exhibit 8 Our Three Scenarios Set the Widest Plausible Band, in Our View, Around Implied Consensus Expectations; They Range from More Bullish Than the IMF's Current Forecast, Through Official Chinese GDP Targets, to a Bear Scenario in Which No Supply Is Removed from the Market Despite Prices Collapsing

Source: IMF and Bernstein estimates and analysis.

The previous sections looked at analysis conducted during 2Q:13. Here, we provide an overview of our updated iron ore price forecast. At the time of publication of this Blackbook, we predict $130/t (2013E) rising to a peak of $144/t (2016E). This is consistent with the IMF's current GDP forecast of 7.8% in 2013 rising to 8.5% in 2016 and average increase in supply of 88Mtpa 2013-16E. By 2020, we expect iron ore prices of $115/t consistent with of IMF forecasts for an average 8.5% Chinese GDP growth 2017-20E and a further 28Mtpa of incremental supply in the market.

Exhibit 9 Our Iron Ore Price Forecast Is Consistent With the IMF's Current GDP Forecast of 7.8% in 2013 Rising to 8.5% in 2016 and Average Increase in Supply of 88Mtpa 2013-2016

Source: Bloomberg L.P. and Bernstein estimates and analysis.

2005A 2006A 2007A 2008A 2009A 2010A 2011A 2012A 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

Consensus 33 42 46 76 81 147 168 128 127 118 110 107 100 95 90 90

Scenario 1: IMF 8.5% 33 42 46 76 81 147 168 128 146 151 162 170 167 160 137 121

Scenario 2: Official Target 7.5% 33 42 46 76 81 147 168 128 137 134 127 125 122 117 106 104

Scenario 3: China Bear 6% 33 42 46 76 81 147 168 128 123 106 95 89 77 68 66 66

-

20

40

60

80

100

120

140

160

180

Iro

n O

re P

ric

e —

US

$/t

Iron Ore Price by GDP Scenario

Nominal Dollars Spot* 2013 YTD 2013E 2014E 2015E 2016E

Iron Ore — US$/t 109 129 130 137 141 144

Iron Ore Consensus — US$/t 123 114 106 100

SCB Upside/Downside 6% 20% 33% 44%

*as of June 30, 2013

Our Current Forecast Is for Iron Ore That Peaks at $144/t in 2016 and Declines to $115/t by 2020



AND LOVE THE ORE

As JK Galbraith observed, "faced with the choice between changing one's mind and proving that there is no need to do so, almost everyone gets busy on the proof." As an exercise in listening to Galbraith, we asked ourselves "what are the valuation impacts if we are wrong, and if $80/t iron ore becomes a near-term reality?" While examining this hypothetical situation, we came across three "automatic stabilizers" that would mitigate a part of the negative value impact of a sustained iron ore price fall for the big Western miners and offer a realistic picture of how a sustained fall in iron ore prices would impact valuations: (1) the natural FX hedge embedded in the miners' cost structures; (2) value-in-use adjustments commanded by ores of differing qualities and (3) elasticity of iron ore demand and spare capacity in China's steel-making industry whose exploitation would become economical, if the iron price were to fall sufficiently. Across all of these stabilizers, those companies that are the most exposed to iron ore experience the greatest offsetting effect.

A natural hedge exists in the operating cost structure of the miners that softens the earnings impact of any sustained fall in commodity prices (see Exhibit 10 and Exhibit 11). In a world of falling commodity prices, we expect to see substantial weakening in producer currencies (AUD and BRL). Given that the miners' costs are denominated in local currency and their revenues in U.S. dollars, any producer-currency weakening will lead to an improvement in operating margin. With nearly 50% of the cost increases in iron ore over the last 10 years coming from the impact of currency movements, any currency weakening is a significant source of upside. We estimate that Vale and Rio would experience the greatest benefits (equivalent to 12% and 8% of June 30, 2013 share prices, respectively), while more diversified BHP and Anglo would experience 4% and neutral impacts, respectively.

Exhibit 10 Mining Costs and Revenue Exhibit a High Degree of Correlation

Exhibit 11 Currency Appreciation Drives a Significant Part of the Cost to Price Relationship, Particularly for the Australian Producers

Source: Bloomberg L.P. and Bernstein analysis. Source: Bloomberg L.P. and Bernstein analysis.

3 Note that we omit Glencore Xstrata from this automatic stabilizer valuation analysis as the company lacks direct iron ore production bar its 1Q:13 acquisition of a stake in Ferrous Resources Ltd. Note that all valuation offsets are expressed in percentage of the June 30, 2013 price in USD terms.

R² = 0.9048

0

5,000

10,000

15,000

20,000

25,000

0 20,000 40,000 60,000

To

tal C

ove

rage

Co

sts

($m

)

Total Coverage Revenue

Total Iron Ore Costs and Revenue

41%

17%

42%

Rio Tinto Iron Ore Cost Drivers

Producer Currency Appreciation Local Currency Inflation

Real Cost Increases

What If We Are Wrong and There Is a Sustained Iron Ore Price Fall (to $80/t)? If So, Three Factors Would Mitigate the Negative Value Impact for the Big Western Miners

We Estimate That Cost Savings, Combined With the Natural FX Hedge Embedded in Operating Costs, Could Offset Valuation Declines by an Equivalent of, on Average, 5% of June 30, 2013 Share Prices for Rio, BHP, Anglo and Vale3



AND LOVE THE ORE 11

Not all iron ores were created equal. Substantial quality and value-in-use (VIU) heterogeneity exists in iron ore products. Removing homogeneity from a commodity industry, of course, has a "de-commoditizing" effect — iron ore grade accounts for 93% of the iron ore price variation around the 62% Fe benchmark. However, even after this effect is adjusted for, poorer-quality, low-grade iron ore still incurs a pricing penalty (see Exhibit 12). Inclusion of a VIU adjustment in the cost curve thus increases the marginal cost of production, reduces the ability for low-quality supply to compete and provides price support for the major miners with high-quality deposits. We estimate that Vale and Rio would experience the greatest benefits (equivalent to 11% and 9% of June 30, 2013 share prices), while more diversified BHP and Anglo would experience 4% and 6% offsets, respectively.

Exhibit 12 There Is a Significant Range of Variation Around the Benchmark 62% CIF Price Point, and Even After Adjusting for Fe Grade, Poor-Quality Ore Suffers a Significant Discount

Exhibit 13 We Believe That a Significant Part of the Slowdown in Chinese Steel Growth Resulted from the Budgetary Constraint Imposed by High Commodity Prices

Source: BAIINFO and Bernstein analysis. Source: Bloomberg L.P., AME, CRU, IMF and Bernstein estimates and analysis.

Chinese steel making has seen its EBITDA margins cut by ~70% over the last decade on the back of rising raw material prices (iron ore CAGR of 26%, PCI and coking coal CAGR of 14%) and essentially stagnant steel prices (local currency CAGR of 1%, USD CAGR of 4%). We believe that this decline stands behind the fall in Chinese steel production growth (down from ~20% p.a. 10 years ago to the ~6% p.a. seen today). Steel capital stock in China is still only at around one-third of the level seen in the West, and the urbanization and industrialization of China remains far from complete (see Exhibit 13). In a world where iron ore starts to fall toward a non-Chinese cost floor of US$80/t, there would be a significant reduction in the cost pressures facing the Chinese steel industry, which, we expect, would yield an increase in the Chinese steel trend growth and push iron ore price back up to US$95/t. Vale and Rio would experience the greatest benefits (equivalent to 31% and 24% of June 30, 2013 share prices), while more diversified BHP and Anglo would experience 10% and 16% offsets, respectively.

R² = 0.7734

-40

-30

-20

-10

0

10

20

30

50% 55% 60% 65% 70%

Iro

n O

re P

rem

ium

vs

. Be

nch

mar

k (U

S$/

t)

Fe Grade of Iron Ore

Price Impact of Quality in Iron Ore

-10%

-5%

0%

5%

10%

15%

20%

25%

30%

35%

0%

10%

20%

30%

40%

50%

60%

70%

80%

Jan-

05

Aug

-05

Mar

-06

Oct

-06

Ma y

-07

Dec

-07

Jul-0

8

Feb-

09

Se p

-09

Apr

-10

Nov

-10

Jun-

11

Jan-

12

Au g

-12

Ste

el G

row

th

Ste

el E

BIT

DA

Mar

gin

Chinese Steel Growth and Profitability Since 2005

Margin, %Annualized Grow thSteel Grow th Ex Monetary Grow th

VIU or Heterogeneity in a Commodity Offers a Degree of Automatic Stabilization to the Iron Ore Price, Offsetting Valuation Declines in an Oversupply World by an Equivalent of, on Average, ~7% of June 30, 2013 Share Prices for Rio, BHP, Anglo and Vale

Chinese Steel Growth Rate Declines Are a Consequence of the Budgetary Constraint of High Iron Ore Prices; Consequently, Low Iron Ore Prices Would See Demand Rise, Offsetting Valuation Declines by an Equivalent of, on Average, 20% of June 30, 2013 Share Prices for Rio, BHP, Anglo and Vale



AND LOVE THE ORE

Mining companies are operationally and financially geared to their underlying commodity exposure, so we provide two valuation metrics (see the "Valuation and Risks" chapter of this Blackbook for more details): ~80% of weekly mining equity price moves can be explained by underlying

commodity price moves, so we use a regression-based trading model and our commodity price forecasts to help determine our 12-month price targets. If the regression remains stable or deviations appear temporary, the model determines the target price. If we believe a deviation is signaling a fundamental change, we will adjust our target price for this fundamental shift and disclose the manner and magnitude of the adjustment made. At present, no adjustments have been made.

We additionally provide a supplementary DCF-based valuation constructed in nominal local currency terms out to 2030 over which explicit commodity price and exchange rate forecasts apply. The nominal local currency cash flows are de-escalated into real U.S. dollar cash flows and discounted at the company-specific WACC. A country risk premium reflecting the geographic origin of the cash flows is added to the underlying WACC to reflect cash flow items (i.e., expropriation) that cannot be explicitly modeled in the cash flow. All reserves are considered exploited by the model. In addition, 50% of the incremental resources (i.e., 50% of the residual resources, excluding those that have already been converted to reserves) of the company are modeled. Where residual life of mine (LOM) may be inferred for operations beyond the 2030 time horizon, a terminal value is applied for the remaining years of potentially exploitable material. We forecast our models in reporting currency (USD), convert to listing currency (GBP or Real), and round final DCF values in 25p/cent increments.

The four most significant risks facing the major mining houses are: 1) lack of capital discipline (specifically displacement of high-cost Chinese marginal producers by low-cost Western production), 2) operating cost inflation (U.S. dollar denominated unit costs in all the major mining houses have seen double-digit growth rates over the last 10 years, roughly half of which are macro related and the other half are real local currency), 3) a sustained downturn in the Chinese economy (the largest consumer of global resources), and 4) resource nationalism (ranging from increased share of rent extraction to outright asset confiscation). For more details on both sector risks and company-specific risks, see the risk section in the "Valuation and Risks" chapter.

We believe that there are sound fundamental reasons why commodity prices over the last five years have been at elevated levels relative to historical precedent. Moreover, as we have discussed, for iron ore4 these reasons are based on supply and demand dynamics that center around China, the largest consumer of commodities. On the demand side, the Chinese government is facing an imperative to industrialize at an unprecedented rate ahead of an equally unprecedented demographic shift. Meanwhile, the Chinese domestic mining industry, with its high-cost and labor-intensive production, has become more important to global supply. We believe that changing the cost structure and importance of this industry may take longer than the market currently anticipates. As a result, we do not believe that the sustained and near-term fall in commodity price implied by mining valuations is a foregone conclusion. Furthermore, in the event that we are wrong, the natural FX hedge embedded in operating costs, VIU adjustment and reduced cost pressures on the steel-making industry that would eventuate would all mitigate the valuation impact of such a fall. A key conclusion of our analysis is the sensitivity of the iron ore price forecast to two variables: China's macro policy (will the outcome be similar to that which the IMF predicts?) and the strategic choices of the incumbent iron ore miners (will they discipline capital in pursuit of value or

4 European Metals & Mining: Iron, Cold Iron, Is Master of Them All...or at Least 60% of EBITDA.

Valuation Methodology

Risks

Investment Conclusion



AND LOVE THE ORE 13

will they simply pursue volume at the cost of suppressing prices?). Both bear watching — and while the former is beyond the control of our coverage companies, the latter component of their destiny is firmly within their own control.

Rio Tinto (TP £41.25; unmodified from trading model prediction) — 73% of 2012 EBIT from iron ore, plus world-class copper exposure: We consider Rio Tinto the most attractive stock in our coverage. Rio owns some of the highest quality iron ore assets and (vitally) infrastructure globally. Rio also has exposure to some of the world's best operational copper assets — not to mention three of the world's best undeveloped copper deposits (Resolution and La Granja). Tier 1 assets in iron ore include Dampier and Cape Lambert (Australian Pilbara), while copper includes Escondida, Grasberg, Bingham Canyon and Oyu Tolgoi. The prospects of genuine capital discipline and cost cutting under new CEO Sam Walsh (who made his bones in low-cost brownfield Pilbara production), coupled with a more reasoned approach to volume growth, mean that we continue to see Rio Tinto as our top pick. Our trading model, used to set 12-month target prices, has shown no signs of the regression shifting (R-squared = 88%) nor is there anything in our fundamental research to date that causes us to anticipate it will, hence we have taken our target price from the unmodified regression model price.

Vale (TP BRL 46.50; unmodified from trading model prediction) — 95% of 2012 EBIT from iron ore, a commodity in which we see near-term upside: Vale is the world's largest iron ore producer and, in Carajas, has one of the most globally attractive iron ore assets. Vale is the most operationally geared miner to the iron ore price and we see an asymmetric risk to the upside in iron ore prices in the near term, hence from a pure value consideration, we are more favorably inclined to Vale now than previously. We do note the significant influence of the Brazilian government (5.5% directly through Golden Shares and ~34% indirectly through the strategic consortium of Valepar). The company is, in our view, most at risk should a reduced iron ore price eventuate, given its geographic longinquity from the world’s largest consumer of iron ore, China. Vale's regression continues to show instability in our trading model (despite R-squared of 86% in our most recent update). However, for now we have taken the unmodified price target generated by the trading model, but we will continue to monitor the regression.

BHP (TP £22.50; unmodified from trading model prediction) — highest-quality stock, Tier 1 assets and limited geographic risk: BHP Billiton not only has the longest corporate history of our coverage group (its component parts trace their origins back to 1851 and 1883, respectively), we view it as our highest quality stock. In 2012, the company generated the highest revenue of the "Big Three" miners in our coverage (US$72 billion versus US$56 billion for Rio and US$49 billion for Vale) and had the highest EBITDA margin (42% versus 39% for Vale and 36% for Rio). BHP Billiton is the second most diversified (from a commodity exposure perspective) company in our coverage, after Anglo American. However, where platinum has been a drag on Anglo's portfolio, BHP's has received a boost from "black gold" in its petroleum division (69% EBITDA margins in calendar 2012 versus 11% for Anglo's platinum division perspective). The strong commodity risk diversification of BHP Billiton is complemented with geographic diversification that, while skewed to Australia, is nonetheless low risk. We have not seen signs of the regression shifting (R-squared =78%) in our trading model, nor is there anything in our fundamental research to date that causes us to anticipate it will. Accordingly, we set our target price from the unmodified regression model price.

Anglo American (TP £20.25; unmodified from trading model prediction) — a soluble turnaround story: Anglo American has the most diversified portfolio of our coverage group from a commodity perspective and is significantly smaller than the "Big Three" of BHP, Rio and Vale due to its smaller iron ore exposure. That Anglo is ahead of Glencore Xstrata on this metric is, in large measure, due to the existence of Kumba in Anglo's portfolio. Kumba, however, like AmPlats and thermal coal, increases the company's country risk exposure due to its South



AND LOVE THE ORE

African location. Anglo American has the highest country risk premium in our coverage — 2.6% versus an average of 1.2% for the "Big Three." Given recent operational difficulties (including failure to deliver on Minas Rio and Los Bronces, not to mention the issues plaguing platinum), Anglo remains in our minds a turnaround story — one that will challenge new CEO Mark Cutifani, but one which we believe is solvable for a leader with his 36 years of operational expertise. Given the shift in the regression and the fundamental factors that have been weighing on the stock, as well as the upcoming 1H reporting (at which time Mr. Cutifani will present the results of his portfolio review), and that we consider Anglo a restructuring story, we consider there is a greater likelihood of regression instability going forward than for Rio or BHP and are monitoring the situation. However, for now, we have taken the unmodified price target generated by the trading model (R-squared = 85%) at present.

Glencore Xstrata (TP £5.25; unmodified from trading model prediction) — traders, copper and coal: Glencore Xstrata offers exposure for copper and coal bulls without a taste for iron.5 It also offers exposure to the sales and trading house that Mr. Ivan Glasenberg built. As such it has a profile that is distinct relative to the pure miners in our coverage. The impact of the high-turnover, low-margin trading business (2% EBITDA margin in 2012 versus 25% for Glencore Xstrata combined) is clearer even when contextualized against the average 33% for our coverage group. As an owner operator (Mr. Glasenberg holds 8% of the combined entity's paper), Mr. Glasenberg's incentives are squarely aligned with investors. It remains to be seen how he will staff the combined entity, following the departure of much of the senior management talent from Xstrata. We do consider that Glencore has some ways to go yet before it will be "institutional quality" on the metrics of reporting and governance. Furthermore, the company's operations in frontier jurisdictions like the DRC not only result in the second-highest country risk premium in our coverage but also carry headline risk — as does Glencore's trading division.6

For Glencore Xstrata, we have constructed a synthetic stock using Xstrata’s actual history and its recent relationship with Glencore. This synthetic shows, in our view, an acceptably high R-squared of 64%. For now, we have taken the unmodified price target of £5.25 but will continue to monitor the trading model.

5 Given that the company's only direct iron ore exposure is its 1Q:13 acquisition of a stake in Ferrous Resources. 6 To wit the April 21, 2013 story Glencore traded with Iranian supplier to nuclear program http://www.guardian.co.uk/business/2013/apr/21/glencore-trade-iran-supplier-nuclear



AND LOVE THE ORE 15

If China Is to Grow Rich Before It Grows Old, It Must Finish Industrializing in the Next Decade

The industrial behemoth,7 that is the Chinese economy, has so far been powered by exploitation of China's most abundant resource: vast pools of surplus rural labor. However, this labor supply is not infinite. Over the next generation, China will experience a substantial increase in its old-age dependency ratio as the country ages sharply in response to earlier policy initiatives (namely, the "one child" policy). A January 2013 IMF working paper8 sees cheap labor in China becoming exhausted sometime between 2020 and 2025 — the same point at which we forecast that China's absolute demand for raw materials will peak. We believe that it is the proximity of this labor exhaustion that has contributed to the urgency of the Chinese government's push to install a productive capital base and transition the economy away from subsistence agriculture to tertiary forms of value-added activity. The unprecedented rapidity of Chinese industrialization has rendered China the largest consumer of global commodities (56% of total iron ore demand globally in 2012); in our view, this reflects the county's requirement to "grow rich before it grows old" and before the supply of cheap labor that has powered this industrialization is exhausted.

It is the rapidity of this industrialization, coupled with the fundamental geological scarcity of certain key raw materials domestically, that has been the key driver behind the emergence of the commodity "super-cycle." In this chapter, we show how our counter-consensus view on "stronger for longer" commodity prices is consistent with the imperative faced by the Chinese to build an industrial society in a little more than 20 years.

Global Demographics and the Impact of China A clear relationship exists between the wealth of societies and their demographic structure (see Exhibit 14). Rich industrialized economies tend to see falling birth rates and rapidly increasing dependency ratios.9 In these societies, old-age provision is not critically dependent on having a large number of children. Rather, the increase in productivity associated with industrialization enables the society to easily generate the output required to support what would otherwise be a significant burden. On the other hand, in societies relying on subsistence agriculture, the birth rate tends to stay high and the society does not age significantly over time.

The transition from an agrarian economy to an industrial society, dominated by tertiary forms of economic activity, requires significant consumption of steel and other raw materials. In effect, it represents the build-up of capital stock that increases the productivity of the workforce, thus liberating labor from the requirement to be engaged in primary enterprises in the process. We see this pattern clearly in the U.S., where the employment in agriculture peaked in 1910 at 11.6 million people before falling monotonically to just over 2 million people today (see Exhibit 15). This is despite the fact that over the same period U.S. production of wheat rose from 17mt in 1910 to over 60mt today (+253%). Likewise, the U.S.

7 The original biblical Behemoth is described as having a tail like cedar, bones like iron, limbs as strong as copper, and sinews like stone, hence it seems an apropos metaphor for the economy of the world's largest consumer of commodities. 8 "Chronicle of a Decline Foretold: Has China Reached the Lewis Turning Point?" IMF WP/13/26. 9 Defined as the number of non-productive workers (according to the UN, those aged 0-14 and 65 plus) relative to the working-age population. So, a dependency ratio of 50 implies 50 non-productive citizens supported by 100 people of working age.

Over the Next Generation, China's Old-Age Dependency Ratio Will Increase Sharply in Response to Earlier Policy Initiatives

In 2000, China's Demographic Composition Resembled That of a Rich Industrial Society, With a Falling Birth Rate and a Rapidly Aging Society, But Its Economy Was Still Agrarian



AND LOVE THE ORE

experienced similar increases in the production of other agricultural commodities (maize +315%; rice +1,970%; cattle +62%). As a result, the service sector employment rose substantially from 5.9 million people in 1910 to roughly 48 million today (see Exhibit 16). In 1850, U.S. agriculture employed 5x as many people as the service sector did (5 million versus 0.9 million), but it now employs only 0.05x as many (2.2 million versus 48 million) — a 100-fold decline over the space of the U.S.'s industrialization.

As the U.S. transitioned from an agrarian to an industrialized society, its capital stock rose. Rather than relying on the energy of human labor to drive output, it is far more productive to harness the forces of mechanization and industrialization (see Exhibit 17). Moreover, this transition is not isolated to the U.S., but has been replicated in one form or another across all countries that have industrialized.

Exhibit 14 There Is a Clear Relationship Between the Aging of a Population and Its Wealth…Usually, Birth Rates Fall as Societies Get Rich, With Aging Population Supported by the Increased Productivity Associated With Industrialization

Source: UN and Bernstein estimates and analysis.

R² = 0.4677

-10

0

10

20

30

40

50

60

-50 -40 -30 -20 -10 0 10

GD

P/C

apit

a —

US

$'00

0


Aging of Population vs. Development of Economy Excluding ChinaSize of Bubble = Population



AND LOVE THE ORE 17

Exhibit 15 In 1910, the U.S. Began Transitioning from an Agricultural Economy…

Exhibit 16 …Toward One Driven by Services

Source: Mitchell and Bernstein analysis. Source: Mitchell and Bernstein analysis. Exhibit 17 We Believe That the Increases in Labor Productivity, Occasioned by the Increased

Intensity of Capital Stock (Machinery, Railroads, etc.), Enabled This Transition

Source: World Steel Association (WSA), UN, Mitchell and Bernstein estimates and analysis.

The stark outlier from the rule of an aging population equating to high output and increased prosperity, however, was China in 2000. In particular, its output in real terms stood at a little over US$2,000 per capita, and yet the country was set to experience a precipitous decline in the working-age population (see Exhibit 18). Little wonder that as few as six years ago, a Chinese demographer reportedly stated at Davos that "China might have to resort to mass suicide in the end, shoving pensioners onto the ice."10 Though stark hyperbole, this comment nevertheless illustrates the realization by the Chinese policy makers that a graying population in a society that has not yet industrialized is a recipe for social and economic disaster. Since 2000, China has embarked upon a program of urbanization unprecedented in human history. Agricultural employment in China peaked in 2000 and has been

10 As reported in the Daily Telegraph on February 6, 2013.

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

1850

1860

1870

1880

1890

1900

1910

1920

1930

1940

1950

1960

1970

1980

1990

2000

2010

Th

ou

san

d P

eop

le

US Agriculture Employment

0

10,000

20,000

30,000

40,000

50,000

60,000

1850

1860

1870

1880

1890

1900

1910

1920

1930

1940

1950

1960

1970

1980

1990

2000

2010

Th

ou

san

d P

eop

le

US Services Employment

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

0

1

2

3

4

5

6

1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Cap

ital

Sto

ck o

f Ste

el —

kg

/Cap

ita

Rat

io P

rim

ary

to T

ert

iary

Jo

bs

Evolution of US Economy to Service Sector

Ratio Primary to Tertiary Employment - USA Steel Capital Stock - USA



AND LOVE THE ORE

declining ever since, with first manufacturing and now services increasing in importance to the overall economy (see Exhibit 19 through Exhibit 22). Indeed, there are now more people employed in the service sector in China than in manufacturing and construction.

Meanwhile, the installation of the capital stock required to support an urban service-based society has continued apace (see Exhibit 23) — and has followed the same trajectory as that of the U.S. and other industrialized countries. There are now as many jobs in services in China as in farming, and the country's capital stock comprises ~4 tons of steel per capita. The U.S. reached the point of equivalence between employment in the service sector to agricultural labor in 1930 accompanied by a capital stock of ~5.5 tons of steel per capita. On this basis, Chinese industrialization has been more productive than that seen in the U.S. Both from the perspective of the labor pool composition and capital stock installation, China today resembles the U.S. of the 1930s (see Exhibit 24). The important difference between the two, however, is the speed of the transition toward an urban society. It took the U.S. 80 years to reduce the importance of agriculture from 5x that of services to parity. The Chinese accomplished this in 30 years.

Exhibit 18 In 2000, China Was a Clear Outlier from the Normal Pattern — the World's Largest Country Was Set to Age Rapidly and Yet Was Still at a Pre-industrial Level of Output


-10

0

10

20

30

40

50

60

-50 -40 -30 -20 -10 0 10

GD

P/C

apit

a —

US

$'00

0



USA

Japan

China

South Korea



AND LOVE THE ORE 19

Exhibit 19 It Was Only in the Year 2000 That China Started Reducing Its Framing Dependence

Exhibit 20 1970s Saw the Rise in Construction and Manufacturing Employment Start…

Source: NBS and Bernstein analysis. Source: NBS and Bernstein analysis. Exhibit 21 …Followed in Short Order by the Service

Sector

Exhibit 22 Today, More People Are Employed in Services Than Manufacturing

Source: NBS and Bernstein analysis. Source: NBS and Bernstein analysis.

0

50

100

150

200

250

300

350

400

Mill

ion

Wo

rker

s

Primary Employment — China

0

50

100

150

200

250

Mill

ion

Wo

rker

s

Secondary Employment — China

0

50

100

150

200

250

300

Mill

ion

Wo

rker

s

Tertiary Employment — China

0

100

200

300

400

500

600

700

800

Mill

ion

Wo

rker

s

Chinese Employment Breakdown

Primary Secondary Tertiary



AND LOVE THE ORE

Exhibit 23 The Pattern in China Is Nearly Identical to That Seen in the U.S.; Capitalization Drives Productivity Increases, Which Then Lead to the Emergence of Services

Source: NBS, WSA and Bernstein estimates and analysis. Exhibit 24 Comparing China to the U.S. Suggests That the Structure of the Chinese Economy

Today Is Similar to the U.S. in the 1930s, Both in Labor and Capital Stock; However, the Transition Has Been Far More Rapid in China

Source: NBS, WSA, Mitchell and Bernstein estimates and analysis.

0

1

2

3

4

5

6

7

8

9

10

1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 20100

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

Rat

io P

rim

ary

to T

ert

iary

Jo

bs

Cap

ital

Sto

ck o

f Ste

el —

kg

/Cap

ita

Evolution of Chinese Economy to Service Sector

Steel Capital Stock - China Ratio Primary to Tertiary Employment - China

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

0

1

2

3

4

5

6

7

8

9

10

1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Cap

ital

Sto

ck o

f Ste

el —

kg

/Cap

ita

Rat

io P

rim

ary

to T

ert

iary

Jo

bs

China and US Evolution Compared

Ratio Primary to Tertiary Employment - USA Ratio Primary to Tertiary Employment - ChinaSteel Capital Stock - USA Steel Capital Stock - China

The Chinese economy iresembles that of the

US in the 1930s



AND LOVE THE ORE 21

Lewis Point and the End of Cheap Labor The Lewis point is defined as the point at which surplus agricultural labor becomes exhausted and industrial wages start to accelerate. This, in turn, lowers the industrial sector productivity, placing a drag on the level of investment in an economy. According to the argument, prior to this point, the pool of rural labor acts as a brake on wages in the industrial sector. This, in turn, increases its profitability, hence supporting a higher level of investment.

So far it would seem that (with the exception of raw materials) China has maintained an exceptionally rapid growth rate with (relatively) low inflation precisely because of this mechanism (or so the IMF would argue). However, the pool of cheap labor in the country is soon to be exhausted. In a January 2013 whitepaper, the IMF marks the point of exhaustion somewhere between 2020 and 2025 (see Exhibit 25). This, therefore, marks the point at which capital stock accumulation in China will come under the greatest pressure. Moreover, this point also coincides with our forecast for the peak demand for raw materials in China, and this coincidence is not accidental (see Exhibit 26). By 2020, the Chinese will have installed a capital stock comparable to that seen in most other industrialized nations. Consequently, China should be able to bear the acceleration in industrial wages with some degree of equanimity — principally because once the country's capital stock is already built, the decline in investment profitability will not matter as much as it would now.

To see this more clearly, let's consider Japan. It started to exhaust its working-age population some 20 years prior to China, with Japanese population in the 15-64 age group growth turning negative in 2000 (see Exhibit 27). However, at this point, Japan was already an industrialized nation with 14 tons of steel per capita in its capital stock.

Exhibit 25 The Demographic Challenges Facing China and the Emergence of Lewis Point Stand

Behind the Requirement to Accelerate the Capital Stock Formation

Note: 2015-40 numbers are UN estimates.


-10%

-5%

0%

5%

10%

15%

20%

1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 2040

Ch

ang

e in

Wo

rkin

g-A

ge

Po

pu

lati

on

China — Growth in Working-Age Population (15-64)

Critical demographic inflection point ~2020 to

2025

As China's Supply of Cheap Rural Labor Is Exhausted, Maintaining the Level of Capital Investment Will Become More Difficult; China's Transition Will Have to Be as Rapid as It Has Been So Far and Continue Until at Least 2020



AND LOVE THE ORE

Exhibit 26 The Urgency to Industrialize the Chinese Economy Is Evident in the Acceleration in Steel Demand in Comparison to the Historical Precedent in the West

Note: 2015-30 numbers are Bernstein estimates.

Source: WSA and Bernstein estimates and analysis. Exhibit 27 A Very Similar Demographic to China Was Seen in Japan Toward the End of the Last

Century; However, by the Time the Japanese Transition Occurred, Industrialization Was Complete and Output Stayed High

Note: 2015-40 numbers are UN estimates.


0

500,000

1,000,000

1,500,000

2,000,000

2,500,000

1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030

Cru

de

Ste

el —

ktGlobal Crude Steel Production

EU (27) Other Europe C.I.S North America South America Africa Middle East

China India Japan South Korea Other Asia Oceania Other

Western post‐war capital

accumulation

Western capital"maintenance"...commodities ex growth

Asian capital accumulation

-10%

-5%

0%

5%

10%

15%

20%

1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 2040

Ch

ang

e in

Wo

rkin

g A

ge

Po

pu

latio

n

China Japan

By 2025, Chinese steel per capita will stand at ~10 tons vs. ~4 today

When Japanhit the demographic ceiling, there was an installed stock of 14 tons per

capita



AND LOVE THE ORE 23

If our demand projections prove accurate, by the time China reaches this demographic tipping point, the country too will have industrialized (critically, it will not have done so if raw material demand goes ex-growth from this point). Consequently, China has a very limited window in which to achieve the necessary changes to the structure of its economy and society, and it is clearly seen in the evolution of China's dependency ratio over time. In the 1950s, the infant mortality rate in China stood at ~200 per 1000, but then fell dramatically over the next 20 years to a level well below 50. However, female fertility levels stayed at roughly the same level (around six births per woman) for a considerable period, thus resulting in an explosive population growth, an increase in life expectancy and a rapid fall in the child dependency ratio following the "one child" policy (see Exhibit 28).

The response of the "one child" policy to the problem subsequently sowed the seeds for an inevitable rise in the old-age dependency ratio (see Exhibit 29). In toto, these two effects generated a demographic "wave" of cheap labor moving through the Chinese economy between 1970 and 2020, as the children of the previous pro-growth policy initiatives matured and entered the workforce (see Exhibit 30).

Given that the first part of the boon of cheap labor was effectively squandered during the Cultural Revolution and subsequent political restructuring, only the period from 2000 to 2020 was left available to take advantage of reorientation of mass cheap rural labor in the Chinese economy. So while a similar demographic profile to China was seen 20 years earlier in Japan, the political situation in the Middle Kingdom and the sheer size of the country led to a radically different outcome (see Exhibit 31).

Further compounding this demographic trend is the accompanying behavioral trend. On a 1Q:13 trip to China, our team interviewed management across industries, including sewing machine and baby goods manufacturing, hospitality, fine dining, education, and mining to gather anecdotes about managing a workforce increasingly composed of only children. These "xiao huangdi" (little emperors) are the sole focus of often six doting relatives (two parents and four grandparents). We collected anecdotes that were similar across different industries: only children were often resistant to shift work or overtime, many expected to be allowed to go home for a lunch break, they would refuse to work if they came down with minor colds, if a family member grew ill they would leave without offering any notice, and if they changed jobs they would also often leave without giving notice. There was also discussion of a seasonal pattern of, if not outright productivity disruption, at least disruption from management's perspective: the month following Chinese New Year (and the payment of bonuses) many workers would shop around for offers of new employment at higher wages (particularly in Shanghai and Beijing). They would take these offers back to their current employer in order to negotiate increased compensation. The employers felt that they had no choice but to raise wages for those workers that had proven themselves to be reliable, as many workers were not reliable. Across all the conversations we had, one common thread emerged: behaviorally, management considered their respective workforces less reliable than they had been 5-10 years in the past. To the extent that our interviews can be extrapolated to the nation at large, this suggests that behavioral shifts in the workforce may further increase the time pressure upon the Chinese government to industrialize beyond that suggested by demographics alone.



AND LOVE THE ORE

Exhibit 28 A Rapid Fall in the Child Dependency Ratio Following the "One Child" Policy…

Exhibit 29 …Will Eventually Be Offset by the Extraordinary Increase in the Old-Age Dependency Ratio

Note: 2020-60 are UN estimates.



Source: UN and Bernstein analysis. Exhibit 30 The End of the Last Century Marked an

Opportunity to Deploy Abundant Labor in China to Develop Its Capital Stock…

Exhibit 31 …Mirroring Japan But Scaling Up More Than 10 Fold (127 Million vs. 1,345 Million People)





0

10

20

30

40

50

60

70

80

Po

pu

lati

on

65+

to 1

5-64

Child Dependency Ratio — China

0

10

20

30

40

50

60

Po

pu

lati

on

0-1

4 to

15-

64

Old-Age Dependency Ratio — China

30

40

50

60

70

80

90

Po

pu

lati

on

0-1

4 an

d 6

5+ to

15-

64

Total Dependency Ratio — China

30

40

50

60

70

80

90

100

Po

pu

lati

on

0-1

4 an

d 6

5+ to

15-

64

China and Japan Aging Compared

China Japan



AND LOVE THE ORE 25

Capital Accumulation and Real Commodity Prices Over 1930-1970, there was a clear link between the trend growth in real commodity prices and the growth in Western capital stock (see Exhibit 32). When the West went ex-growth for raw materials following the oil shocks of the 1970s, real commodity prices started their downward trend. This illustrates that falling real commodity prices do not constitute the norm. In fact, quite the opposite should be true, given that the miners start by accessing the most promising and easily exploited geology and move onto the harder, higher-cost deposits only subsequently. However, after the mid-1970s, productivity improvements flattened the mining cost curve. Given the commoditized nature of the industry and absent meaningful marketing power, the benefits of these cost improvements were transferred to raw material consumers through lower prices. Mining companies became their own worst enemies as they competed on volume and pursued simple "beggar thy neighbour" strategies.

This trend halted abruptly in 2003, as the world — driven by Chinese industrialization — re-entered the mode of capital accumulation. Part of our conviction in continued commodity price strength lies in the real and inevitable requirement we see to transform the Chinese economy. We believe that capital accumulation in China must continue to run for some time to come, and therefore demand for raw materials going ex-growth simply does not match the demographic reality faced by the Chinese policy makers.

As an example of this, we show our forecast for real copper prices and the continuation of capital stock accumulation in Exhibit 33. The market currently prices in significant declines in the real copper price going forward. In our view, this is at odds with the historical precedents of significant size economies trying to industrialize and the effect of such industrialization on commodity prices.

A further complication relates to the fact that it has been Chinese mining wage escalation that has driven the commodity "super-cycle" through its impact on the structure of the global cost curve (see Exhibit 34). As China approaches the Lewis point, Chinese wages should start to accelerate. If China has not been able to "unlock" Africa by the time this happens, and absent massive value destructive capital profligacy from the major miners, the chance of seeing further significant commodity price rises cannot be discounted.

The Period of Capital Accumulation in the West Saw a Trend of Increasing Real Commodity Prices; We See Chinese Capital Accumulation Continuing to at Least 2020



AND LOVE THE ORE

Exhibit 32 1930-1973 Saw Real Commodity Prices Grow as Capital Was Accumulated in Western Economies; Once Demand Went Flat, Productivity Improvements in Mining Drove Real Commodity Prices Down

Source: WSA, Bloomberg L.P., Mitchell and Bernstein estimates and analysis. Exhibit 33 The Emergence of a New Phase of Capital Accumulation Driven by China Has Seen

Significant Real Commodity Price Appreciation; Our Counter-Consensus View on Commodities Is Driven, in Part, by a Belief That This Will Continue Until the Chinese Capital Stock Accumulation Comes to an End

Note: 2015-20 are Bernstein estimates

Source: WSA, Bloomberg L.P., Mitchell and Bernstein estimates and analysis.

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

2,000

3,500

5,000

6,500

8,000

9,500

11,000

12,500

1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000

Rea

l Co

pp

er P

ric

e —

US

$/l

b

US

Cap

ita

l Sto

ck—

kg

Ste

el/C

apit

a

Real Commodity Prices and Capital Stock Accumulation to 2003

US Capital Stock Real Copper Price

In an ex. demand growth world, productivity

improvements drive down real commodity prices

0

1,500

3,000

4,500

6,000

7,500

9,000

10,500

12,000

13,500

1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 20200.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

Cap

ita

l Sto

ck —

kg S

teel

/Cap

ita

Rea

l Co

pp

er P

rice

—U

S$

/lb

The Rise of China and the Impact of Real Commodity Prices

Real Cu Price Real Cu Price Forecast US Capital Stock China Capital Stock



AND LOVE THE ORE 27

Exhibit 34 Chinese Labor Costs Have Been Instrumental in Changing Global Commodity Cost Curve Structures and, With Them, the Price of Mined Raw Materials; What Happens Once China Moves Past the Lewis Point?

Source: IMF, NBS and Bernstein estimates and analysis.

0

50

100

150

200

250

300

350

400

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

Chinese Costs and Global Growth vs. Price

IMF Metals Index Chinese Costs + GDP



AND LOVE THE ORE



AND LOVE THE ORE 29

Iron Ore "Super-Cycle" Pricing Generated by High-Cost Producers Responding to the Rapidity of China's Industrialization

In our 2012 Blackbook, European Metals & Mining: Iron, Cold Iron, Is Master of Them All...or at Least 60% of EBITDA, we reviewed why iron ore is important to our coverage group, and discussed in length our supply and demand framework.11 Iron ore is the key commodity for our coverage group, generating, on average, 41% of revenue and 65% of EBITDA in 2012. This has risen from the average of 23% of revenue and 28% of EBITDA back in 2003, supported not only by rising iron ore prices12 but also by iron ore production increases of 11.6% CAGR across our coverage group.13 In our view, it is the urgency of China's industrialization (discussed in the previous chapter), coupled with the fundamental geological scarcity of certain key raw materials domestically, that has been the key driver behind the emergence of the commodity "super-cycle." On the demand side for iron ore, our secular commodity price view is supported by a fundamental analysis of industrialization patterns in 120 countries since 1900. On the supply side, we identify the emergence of a new class of marginal iron ore producers in 2003 (Chinese high-cost producers), their impact on the global cost curve, and the duration of supply side support for "super-cycle" pricing based on new greenfield and brownfield projects. We see the "super-cycle" pricing regime supported on the demand side until China finishes industrializing (2020-25) and on the supply side until sufficient new low-cost Western production is brought on line to displace the marginal high-cost Chinese producers (the timing of which rests in the hands of the Western majors and underlies our continued — and continuous — calls for capital discipline).

Iron Ore — Key Commodity for Our Coverage Iron ore is the key commodity for our coverage group, generating, on average, 41% of revenue and 65% of EBITDA in 2012 (see Exhibit 35 and Exhibit 36). This has risen from the average of 23% of revenue and 28% of EBITDA back in 2003, supported not only by rising iron ore prices but also by iron ore production increases of 11.6% CAGR across our coverage group. Of the "Big Three,"14 BHP showed the highest CAGR for the period (8.1%), while Vale — which started from a production base 2.5x that of BHP's in 2003 — had the lowest CAGR at 6.9% (meaning that by 2012, Vale's production base was only 1.9x that of BHP's). Anglo American had the highest CAGR out of our coverage stocks exposed to direct iron ore production of 23.7% (see Exhibit 37 through Exhibit 40).

11 Please see the "China Growth Scenarios" and "Our Iron Ore Price Forecast" chapters for more information on our price forecast in the context of different scenarios discussed in this Blackbook. 12 2003-12 saw the IMF 62% FE monthly spot (CFR Tianjin port) rise at a CAGR of 128%, from an average $14/t to an average $129/t. 13 Excluding Glencore, which has no direct production exposure to iron ore, bar the company's 1Q:13 stake acquired in Ferrous Resources. 14 Rio Tinto, Vale and BHP Billiton.

We See Iron Ore "Super-Cycle" Pricing Continuing Until High-Cost Chinese Marginal Producers Are Displaced by New Low-Cost Production

Iron Ore’s Contribution to Our Sector’s EBITDA Has Increased from 28% to 65% (2003-12), Supported by a Price CAGR of 128% and Production Volume CAGR of 12% Over the Period



AND LOVE THE ORE

Exhibit 35 Across Our Coverage Group, Iron Ore Generated, on Average, 41% of Revenue…

Exhibit 36 ...And 65% of EBITDA in 2012

Source: Corporate reports and Bernstein analysis. Source: Corporate reports and Bernstein analysis. Exhibit 37 Iron Ore Currently Accounts for 44% of Rio Tinto's Revenue and Has Been Driven by

a 7.6% Production CAGR

Source: Corporate reports and Bernstein estimates and analysis.

70%

44%

31%

18%

41%

2012 Revenue from Iron Ore

97%

76%

49%

37%

65%

2012 EBITDA from Iron Ore

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

Re

ven

ue

—$m

Pro

du

ctio

n —

kt

Rio Tinto Iron Ore

Rio Tinto Volume Rio Tinto Revenue



AND LOVE THE ORE 31

Exhibit 38 Iron Ore Currently Accounts for 31% of BHP Billiton's Revenue and Has Been Driven by an 8.1% Production CAGR

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 39 Iron Ore Currently Accounts for 70% of Vale's Revenue and Has Been Driven by a

6.9% Production CAGR


0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

80,000

90,000

Re

ven

ue

—$m

Pro

du

ctio

n —

kt

BHP Iron Ore

BHP Volume BHP Revenue

0

5,000

10,000

15,000

20,000

25,000

30,000

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

160,000

180,000

Rev

enu

e -$

m

Pro

du

ctio

n -

kt

Vale Iron Ore

Vale Volume Vale Revenue



AND LOVE THE ORE

Exhibit 40 Iron Ore Currently Accounts for 18% of Anglo American's Revenue and Has Been Driven by a 23.7% Production CAGR


Iron Ore Demand Iron ore demand can be understood only in the context of steel demand, as steel is the means by which iron ore is actually consumed in an economy. Through months of data analysis, we have assembled a database covering the consumption of metals and economic output for 120 countries since 1900. Our analysis identified distinct and repeatable patterns in every country that has successfully industrialized (as well as the anomalous patterns in those that have not).

First and foremost, it is important to distinguish between the rate of steel consumption (how much is being consumed/embedded today) and the level of steel consumption (how much has been consumed/embedded historically). We liken the situation to filling a bathtub: the speed of water coming out of the tap cannot tell you when the bathtub will overflow — you also need to know how full the bathtub is already and how much water it can hold.

Industrializing and industrialized economies move through three distinct phases in their rate of steel consumption. The same three phases are repeated across industrialized and industrializing countries in our dataset.

i. Development. The build-up of industrialization where steel use grows faster than the underlying GDP. This corresponds to the phase of capital-base installation, which will subsequently generate output.

ii. Peak. A peak in intensity where GDP growth and steel growth are matched. During this phase, we see the effect of diminishing returns — additional capital spending starts delivering incrementally less output.

iii. Decline. The development of an economy based on tertiary forms of value-add, i.e., forms of manufacturing that require little incremental consumption of natural resources (e.g., airbags in cars — no new steel is used but significant value is created). Once this occurs, overall economic growth will be faster than the growth in metal consumption.

As a country industrializes, it evolves through these three distinct phases, which result in a unique and consistently repeated structural form that appears across the development of Western Europe, North America and the developed Asian economies. Following the same three-phase trajectory for the evolution of

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

0

5,000

10,000

15,000

20,000

25,000

30,000

Re

ven

ue

—$m

Pro

du

ctio

n —

kt

Anglo Iron Ore

Anglo Volume Anglo Revenue

Our Dataset of 120 Countries Since 1900 Shows All Industrialized and Industrializing Countries Pass Through Distinct and Repeated Patterns of Steel Demand



AND LOVE THE ORE 33

the steel consumption rate within all industrialized and industrializing countries, China's steel intensity has passed from development through peak and is now in the decline phase — a period in which the rate of demand growth is decelerating. The requirement to continue to build the capital stock and complete the industrialization constrains how rapid the deceleration will be. The steady-state rate of consumption, once the process is completed, will be determined by whether the economy, once mature, adopts a more manufacturing or services-oriented focus (see Exhibit 41).15 Countries that have not industrialized (and historical data for industrialized countries prior to industrialization) show clustered and random (albeit typically low) rates of steel intensity.

In addition, three paradigms emerge for the evolution of the level of steel consumption across countries that have industrialized (or attempted to industrialize). These patterns show how the level of steel stock has varied by the stage of economic development in each country, which allows us to view how a country, becoming ever more productive (i.e., increasing total factor productivity with increasing economic development), looks through the lens of cumulative steel use. We dub these paradigms "Services and Technology" (exemplified by the U.S. and the U.K.), "Manufacturing" (exemplified by Germany and Japan) and "Inefficient" (exemplified by the former USSR)16 — see Exhibit 42.

Our framework shows that Chinese steel demand (46% of 2012 finished steel demand globally) is on a sustainable path (see Exhibit 43). We believe it is erroneous to conflate decelerating metal intensity with negative or flat metal demand growth. It is likewise erroneous to conclude that rapidity of development is an appropriate proxy for efficiency (or lack thereof) in capital allocation. However, these seem to be two of the more widely disseminated flaws in analyzing the Chinese steel industry. China’s current steel intensity (rate) is consistent with its state of industrialization and shows that the country is transitioning from Phase 2 (Peak) to Phase 3 (Decline) at a rate commensurate with that seen elsewhere. This is not a "structural shift" so much as a natural development to a state that is less steel intensive. The current trajectory sees China ending up less steel intensive than "Manufacturing" countries like South Korea but more so than "Services and Technology" countries like the U.S. To highlight the sustainability of China's demand, we would call attention to the fact that the economic activity of every person in the U.S. (or any other developed country) is supported by an installed capital base comprising 80kg of copper and 12,000kg of steel — more than triple the current capital base in China (26kg of copper and 4,500kg of steel per capita) — see Exhibit 44.

15 In order to render these comparable across differently populated countries, we adjust output and embedded stock on a per-capita basis.

16 As with the rate calculation, in order to render these comparable across differently populated countries, we adjust output and steel consumption on a per-capita basis.



AND LOVE THE ORE

Exhibit 41 Like Other Countries That Have Industrialized, China's Steel Intensity Has Passed from "Development" Through "Peak" and Is Now in "Decline"; the Steady-State Rate of Consumption Will Be Determined by Whether the Economy, Once Mature, Adopts a More Manufacturing or Services-Oriented Focus

Source: WSA, IMF, Mitchell and Bernstein estimates and analysis.

Exhibit 42 The "Services and Technology" Pattern (Embodied by the U.S.) Shows the Highest Output/Capital Stock (on a Per-Capita Basis), the "Manufacturing" Pattern (Embodied by Germany) Also Shows an Upward-Sloping Relationship Between Output and Capital Stock; Only the Failed "Inefficient" Pattern (Embodied by the former USSR) Shows a Flat and Negative Relationship Between Output and Embedded Capital Stock


0

10

20

30

40

50

60

70

80

0 5 10 15 20 25 30 35 40 45 50

Ste

el In

ten

sity

(kg

/$'0

00

GD

P)

Output — GDP/Capita — $'000 (Real 2005 PPP)

Trend Steel Intensity Evolution

France USA China Japan South Korea China Actual

0

5

10

15

20

25

30

35

40

45

50

0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 20,000 22,000 24,000 26,000 28,000 30,000 32,000

Ou

tpu

t —

GD

P p

er C

apit

a ($

1,00

0)

Capital Stock — kg Steel per Capita

Paradigms for Embedding Steel Stock

USA USSR Germany



AND LOVE THE ORE 35

Exhibit 43 Our Analysis Derives a Consistent Relationship Between Steel Demand Growth and Overall Level of Economic Activity for China; It Does Not See Signs That the Output Generated by the Embedded Capital Stock Is "Inefficient"; Rather, It Is Increasing in a Manner Consistent With Development Into an Economy Midway Between the "Services and Technology" (Embodied by the U.S.) and the "Manufacturing" (Embodied by Germany) Patterns


0

5

10

15

20

25

30

35

40

45

50

0 2,000 4,000 6,000 8,000 10,000 12,000 14,000

GD

P p

er

Cap

ita

—U

S$'

000

Installed Steel Stock (including depreciation) — kg per Capita

Steel Stock vs. Output Evolution

China Historical China Forecast - SCB USA Historical Linear (China Historical - Trend)



AND LOVE THE ORE

Exhibit 44 A Central Plank of Our Thesis Rests on the Fact That Chinese Urbanization and Industrialization Are Still Far from Complete; in Terms of Capital Stock Accumulation, Labor Composition and Levels of Output, China Today Resembles the U.S. in the 1930s


R² = 0.6587

0

10

20

30

40

50

60

70

80

0 5,000 10,000 15,000 20,000

Ou

tpu

t —

GD

P/C

apit

a $'

000

(Rea

l 20

05 P

PP

$)

Capital Stock — kg Steel/Capita (including depreciation)

Capital Stock to Output by Country

China

UK

Kuwait

Germany

S Korea

Taiwan

USA



AND LOVE THE ORE 37

Iron Ore Supply We do not believe that the microeconomics of supply and demand have suddenly ceased to determine commodity prices. In other words, we reject "financialization" or other non-economic explanations of the "super-cycle." Rather, a shift in the source of the marginal unit of supply and lack of transparency into that marginal unit have obfuscated the underlying supply and demand dynamic. In 2003, Chinese producers emerged as the marginal high-cost supplier for metals, including iron ore, met coal and copper. However, a lack of visibility into the nature of these producers led to a breakdown in the representativeness of the dataset historically used to construct the global cost curve (see Exhibit 45 and Exhibit 46). We see a paradox in China: pre-industrial modes of natural resource production coexisting with industrial modes of production in other sectors. This has led to Chinese mining costs outstripping productivity gains, which has driven up both Chinese commodity imports and global commodity prices. Chinese labor productivity and costs and, specifically, the inability to displace increasingly expensive labor with capital represent the true cause of commodity "super-cycle" pricing. Our analysis has completed the global cost curve, showing that ~10% of global iron ore production in 2011 came from the "missing" Chinese mines (see Exhibit 47). From the supply side, the current "super-cycle" is supported so long as these high-cost marginal producers in China are not displaced. China lacks the geological endowment, experience in capital allocation processes and political will to drive a transition to capital-intensive modes of mining. That leaves, in our view, two routes for displacement — Chinese mining abroad (e.g., in Africa) and new low-cost supply brought on by Western majors. This means that the persistence of the supply side support for the "super-cycle" will be determined by the degree of capital discipline exercised by Western majors (see Exhibit 48). Currently, our medium-term iron ore price forecast is based on a "zero-capital discipline" assumption. That is, we assume that the miners will continue failing to prioritize their total portfolio value ahead of the marginal value of incremental growth. To the extent that Western majors bring on new, low-cost supply in copious amounts (referring here to the mega projects that have the ability to displace significant numbers of high-cost marginal suppliers in China), they effectively subsidize Chinese industrialization at the expense of future sector profitability. Thus, the fate of sector profitability rests squarely on the shoulders of the incumbents, who control the best portion of the world's production and new mega projects.

In 2003, Chinese Producers Emerged as the Marginal, High-Cost Suppliers of Iron Ore; But Lack of Visibility Into the Nature of These Producers Led to a Breakdown in the Representativeness of the Dataset Historically Used to Construct the Global Cost Curve — Our Reconstructed Global Cost Curve Includes the "Missing" Chinese Mines



AND LOVE THE ORE

Exhibit 45 At First Glance, the Old Heuristic of Marginal Cash Costs Appears to Have Broken Down…

Exhibit 46 …But the Fact That Our Visibility Into Costs in China Is Incomplete Provides a Better Answer

Source: AME, Bloomberg L.P. and Bernstein analysis. Source: AME and Bernstein analysis.

Exhibit 47 China Accounted for ~16% of 2012 Global Iron Ore Production (ROE Equivalent); Our Analysis, Which Completes the Global Cost Curve, Shows That ~60% of the Chinese Production Consisted Primarily of "Missing" Chinese High-Cost Mines

Exhibit 48 Our Analysis of the Cash Costs of 149 Green and Brownfield Projects Shows That the Majority Will Not Be Value Accretive at Iron Ore Prices Below $150/t Against a Fully Loaded Discount Rate

Note: There are a few other high-cost producers in the "missing Chinese mines" category but they are de minimis.

Source: AME and Bernstein estimates and analysis.

Source: AME, CRU and Bernstein analysis.

0

20

40

60

80

100

120

140

160

180

200

2005 2006 2007 2008 2009 2010 2011 2012

US

$/t

Iron Ore Costs and Price

C(90) Price

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Australia China

Percentage of "Costed" Iron Ore Supply

Known Unknown

170

160

150

140

130

120

110

100

80

70

0

60

180

40

30

20

10

0

50

1,000,000750,000500,000250,000

90

1,250,000


Missing Chinese

Mines

800,000600,000400,000200,0000

5,000

25,000

35,000

45,000

55,000

60,000

20,000

1,000,000

50,000

40,000

15,000

0

30,000

10,000

There are 16 projects globally that need to be understood/monitored to track the future of the iron

ore price. Capital discipline here is key to ensure that the market

remains well supported.

Cumulative production from new Greenfield projects – kt Fe

Pro

duc

tion-

kt F

e



AND LOVE THE ORE 39

Long-Term Price Forecast We see the "super-cycle" pricing regime supported on the demand side until China finishes industrializing (2020-25), and on the supply side, until sufficient new low-cost Western production is brought on line to displace the marginal, high-cost Chinese producers (the timing of which rests in the hands of the Western majors and underlies our continued — and continuous — calls for capital discipline).

We believe that prices rise to the extent that marginal units of supply are called by demand and that prices fall to the extent that marginal units of supply are displaced. Because marginal supply determines price, the cost structure of an industry determines both the revenue and the margin generation of that industry. With this heuristic in place, we take inspiration from Fischer Black's work on business cycles and monetary policy17 and construct a modified framework, replicating his approach to business cycles but extending the lead time between demand signals and the supply side response. One consequence of the long lead time is inherent cyclicality in the mining industry. Furthermore, this cyclicality ensures that the incentive price (the price at which a miner is incentivized to bring on new capacity) is seldom achieved, but is rather over- or under-shot. This means that the value generated by investment (either by the mining companies themselves or by investors into mining companies) is almost entirely a call on the ability to act counter-cyclically.

Our supply and demand analysis gives rise to our counter-consensus positive iron ore price forecast (2013: +3%, 2014: +18% and 2015: +29%) — see Exhibit 49 and Exhibit 50. We expect iron ore to peak at $145/t in 2016 as supported by the marginal costs of Chinese iron ore production. We believe that the tendency to embed backwardation into commodity price forecasts and instantaneous reversion to some notional long-term mean has been the most obvious error in commodity price forecasting over the last 10 years. In our view, it is the supply side cost structure that determines the structural shape of commodity prices over the longer term; supply is sticky as projects have multi-year lead times, while demand gains and losses can be relatively instantaneous. All of which is another way of bringing us back to a principal tenet of this Blackbook: capital discipline is the key to price preservation.

Exhibit 49 We Embed a Contango Rather Than Backwardation Into Our Short- to Medium-Term Price Forecasts

Source: Bloomberg L.P., IHS Global Insight and Bernstein estimates and analysis.

Exhibit 50 Our Iron Ore Forecast Sees Considerable Upside in 2014-16E Relative to Consensus; We Would Be Even More Bullish If We Were Convinced the Western Majors Were Committed to a Higher Degree of Capital Discipline

* As of June 30, 2013


17 "Business Cycles and Equilibrium," published in 1987.

Iron Ore Fines FOB Aus 2013 H2 2014 H1 2014 H2 2015 H1 2015 H2 2016 H1 2016 H2 2017 H1 2017 H2 2018 H1 2018 H2

Nominal Forecast 131 136 139 140 142 143 145 141 146 137 134

Real (2013$) Forecast 130 134 136 135 136 136 136 129 126 123 120


Iron Ore — US$/t 109 129 130 137 141 144



We Apply a Framework Taken from Fischer Black's Work on Business Cycles to Our Supply and Demand Analysis, Adapted to Long Lead Times in the Mining Sector; This Generates Our Counter-Consensus Positive View on Iron Ore Pricing



AND LOVE THE ORE



AND LOVE THE ORE 41

Two Times the Rate at One-Fourth the Labor: Chinese Steel Intensity Does Not Indicate Overinvestment

This chapter grew out of an analysis of productivity increases in mining, which we looked at to address frequent client concerns about the high rate of Chinese steel consumption in recent years. This rate, in both absolute magnitude and relative to the current consumption by other countries, is often interpreted to suggest overinvestment and that the Chinese economy may be unduly steel intensive — in which case, an immediate and sustained iron ore price decline to US$80/t might indeed be reasonable.18 Is China's steel intensity (annual rate of consumption) higher than that of the U.S.? Yes, of course, but...once one adjusts for relative labor productivity in iron ore mining, the answer becomes emphatically “no.” When the U.S. industrialized, steel and bulk commodities traded in local markets. In contrast, China's industrialization reflects a world of global bulk commodity markets, where the productivity of the world's best iron ore miners (Australians) — rather than just the geographically proximate producers — can be utilized to advance capital stock formation. In effect, China has been able to outsource its mining industry to Australia, and in doing so it has increased the effective productivity of its raw material consumption radically. Every million tons of steel embedded into the U.S. economy has consumed 475 man-years of labor in the form of iron ore. On the other hand, every million tons of steel embedded into China (so far) has consumed, on average, only 183 man-years of labor. So while China is embedding steel into its capital stock at approximately double the average speed of the U.S., the steel consumed reflects a use of resources (labor) only one-fourth as intense as that used during the U.S. industrialization. The wholesale acquisition of Western steel plants and an ability to replicate Western levels of steep productivity through the purchase, disassembly and reassembly of steel-making production in China enabled rapid Chinese steel growth (e.g., the Famous Industrials Group industry relocation projects). In essence, China combined the efficiency of large blast furnaces with cheap labor and raw materials to affect an incredibly high return to capital, which has enabled the rapidity of capital stock accumulation in China so far. Consequently, from a demand perspective, we consider the fear that the iron ore price will decline to US$80/t and remain there is overdone.

Steel Intensity and Productivity It is estimated that it took 30,000 people 30 years to build the Great Pyramid of Cheops — a total of 900,000 man-years of labor embedded into 6 million tons of Egyptian limestone (of course, today many would point out that the limestone intensity of Egypt was far higher than that seen in any previous period of temple building. Consequently, they would criticize the endeavor as being illustrative of massive capital misallocation and thereby predict the imminent collapse of Egyptian civilization some 3,000 years too early). The vast increases in labor productivity can be powerfully illustrated by pointing out that the same edifice today could be pulled together by 150 Australian iron ore miners in a year. This

18 In the "What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities?" chapter, we demonstrate that the market is currently factoring in a long-term iron ore price of US$80/t or a 40% discount to the June 12, 2013 spot price.

Every Million Tons of Steel Embedded Into the U.S. Economy Has Consumed 475 Man-Years of Labor in the Form of Iron Ore Versus 183 in China; Absent Western Capital Ill-Discipline, We See Price Support for Iron Ore Until 2023

Neither Farmers Nor Hairdressers Consume a Huge Amount of Metal; Rather, Metal Is Required to Embed the Capital Stock That Allows the Transition from Primary to Tertiary Forms of Value-Add to Take Place



AND LOVE THE ORE

represents a 6,000-fold increase in the efficiency of human labor. We would understand this increase as being driven by the ability of capital stock (machinery) to convert high-intensity energy sources (oil) into useful work far more effectively than the human body can convert food into output. It is against this backdrop that we illustrate the change in the operating model of the Australian mining industry over the last 50 years in Exhibit 51 and Exhibit 52. These exhibits not only illustrate the transition in the mining industry, but also offer a qualitative impression of the main theme of this chapter — the impact that mining productivity improvements have had on the rapidity of capitalization and urbanization in China over the last decade. It is easy to forget how phenomenally effective Rio Tinto, BHP Billiton and the like are at the task of mining (of course, whether their economic analysis is of the same caliber is a different question). The engineering competence on display at well-run mining operations and the increase in this over time have significant implications on how we should understand the development of industrial societies.

Exhibit 53 illustrates the evolution of U.S. industrialization from 1900 to date, exemplified by the rate at which metal is embedded into the economy. The picture is of an agrarian society moving to an industrial society as its economy transitions from one dominated by farming to one that is services focused. At either end of this transition, metal intensity (rate of consumption/embedding) is low: neither farmers nor hairdressers consume a huge amount of metal. Rather, metal is required to embed the capital stock that enables the transition from primary to tertiary forms of value-add to take place.

In this transition, labor productivity plays two vital roles. First, improved agricultural productivity is required to generate a surplus level of output that frees labor from the task of feeding itself and enables it to branch out into other economic activities. The greater the agricultural productivity, the more labor can be freed from farming and moved into the production of raw materials. Next, the greater the level of productivity in mining (for example), the greater the output of raw materials that can be generated by this now-surplus, non-agricultural labor. The greater the output of these raw materials, the greater the rapidity with which a capital stock can be installed. And, of course, the faster the capital stock can be installed in a country (i.e., bridge, buildings, rolling stock, machinery, etc.), the faster the transition to a service-sector economy can be achieved. This is a relatively simple narrative, but one which we have seen repeated over and over in the database we built, which looks at the rates of metal consumption/embedding across 120 countries since 1900.19 Furthermore, we have the same narrative in the back of our minds when considering the current and future role for metals in the global economy.

19 Iron - Cold Iron - Is Master of Them All...(Or At Least 60% of EBITDA) - The Forms of Steel Use & Iron Ore Demand (2/5).



AND LOVE THE ORE 43

Exhibit 51 Mining in Western Australia in 1961...

Source: 1961 Report of the Department of Mines, Western Australia, Presented to Both Houses of Parliament in 1962.

Exhibit 52 ...And Today

Source: BHP Billiton.



AND LOVE THE ORE

Exhibit 53 At Either End of the Industrialization Process, Metal Intensity (Rate of Consumption/Embedding) Is Low; Rather, Metal Is Required to Embed the Capital Stock That Enables the Transition from Primary to Tertiary Forms of Value-Add (or Farmers to Hairdressers)

Source: WSA, Mitchell, IMF and Bernstein estimates and analysis.

Exhibit 54 shows China's path of steel intensity and summarizes why there is significant anxiety about the steel-intensive nature of Chinese economic development. The peak steel intensity in China was ~65kg/$'000 of GDP — some 60% higher than that in the U.S. It should be recalled, however, that in the U.S. steel intensity peaked ~100% higher than the peak steel in France. Consequently, the dangers of drawing uncritical parallels between China and the U.S. are obvious, and yet the fact that Chinese metal intensity is high has led many to fear that it is "unsustainably" high and that a rapid retrenchment from current levels is imminent.

China's overall economic growth is decelerating. Moreover, the country's metal intensity (rate of consumption/embedding) will track below rather than above Chinese GDP growth. However, part of the reason for the decline in metal growth (and thereby indirectly GDP growth) in China is the impact of high commodity prices. To the extent that the world was willing to supply the Chinese economy with nearly "free" raw materials (i.e., iron ore at US$12/t), it showed itself quite able to absorb these raw materials at an unprecedentedly rapid rate. Clearly, there could be hysteresis in this relationship, but there is at least a prima facia case that lowering commodity prices will see Chinese growth re-accelerate.

That China's apparent steel intensity exceeds that seen in the U.S. belies the fact that far more real resources were embedded in the U.S. capital stock than are being embedded in China today. This comes down to the creation of a globalized commodity market and access to 2.5-5x more efficient (in man-hour terms) iron ore mining. This is made possible not only by geological and population differences, but also due to the capitalization of the mining industry. The West underwrote its own industrialization (and that of post-World War II Japan) through the removal of labor from the mining sector in favor of capital. We question how much more Western capital will be embedded into the sector. In the case of the iron-ore mega projects, capital spend represents a value transfer from Western shareholders to

0

5

10

15

20

25

30

35

40

45

50

0 5 10 15 20 25 30 35 40 45 50 55

Ste

el In

ten

sity

(kg

/$'0

00

GD

P)

GDP/Capita (US$'000)

U.S. Finished Steel Demand Intensity Since 1900

SI Actual SI Trend

2012

1900

1917

1975

1948

The Peak Steel Intensity in China's Ongoing Industrialization Was Some 60% Higher Than That Seen in the U.S.; But the Peak Only Indicates the Rapidity of the Industrialization Process —the U.S Peaked 100% Higher Than France When the Latter Industrialized



AND LOVE THE ORE 45

China through suppressing iron ore prices and thereby underwriting the continued industrialization of the Middle Kingdom.20

Exhibit 54 The Chinese Economy Has Seen Steel Intensity Peak at ~65kg/$'000 of Output (vs. 40kg/$000 U.S. Peak); Because China's Peak Was More Than 60% Higher Than the U.S., and Because Many Datasets Look at Only the Last ~20 Years of Data, This Has Spawned Fears That the High Rate of Steel Consumption in Recent Years Is Indicative of Overinvestment and That the Chinese Economy Is Unduly (and Inefficiently) Steel Intensive

Source: WSA, Mitchell, IMF and Bernstein estimates and analysis.

20 The literal translation of Zhong-guo — the Chinese name for China, which can also be rendered as Middle State. America is Meiguo, or Beautiful Kingdom/State.

0

10

20

30

40

50

60

70

80

0 5 10 15 20 25 30 35 40 45 50

Ste

el In

ten

sity

(kg

/$'0

00

GD

P)






AND LOVE THE ORE

Globalization, Freight and China's Industrialization During its industrialization, China has had access to Australian iron ore production (20,000 tons/man-year at peak versus the U.S. peak of ~9,000 tons/man-year). It is not capitalization that made the difference between the U.S. and Australian industries, as both countries have well-capitalized mining industries. Rather, three factors stand behind this higher Australian productivity and China’s access to it: globalization of the commodities industry, different geological endowment and different population densities.

Exhibit 55 and Exhibit 56 display the history of global freight rates, going back to 1750. This is a vitally important place to start when considering the history of bulk commodities economics as freight rates are the driving force behind the decline in regional commodity markets and the rise of a truly global trade in iron ore and coal. Between 1965 and 2005, real freight costs fell 40%, while global trade in iron ore rose 330% and coal 1,070%. Global markets arise when the value of freight is lower than the value of the commodity being traded. These markets fragment again, of course, when freight rates rise relative to the cost of the commodity.21 Prior to the advent of global trade in the bulk industrial commodities (and it is these commodities that form the bedrock of industrial development), the level of national raw material productivity was the critical determinant of the pace of industrial development. In Exhibit 57 through Exhibit 59, we summarize the economic history of the U.S. iron ore industry since 1900. Productivity in U.S. iron ore production peaked at nearly 9,000 tons/man-year, and the path to that level was by no means steady. For large parts of its history, the U.S. was producing iron ore at productivities of less than 4,000 tons/man-year.

Exhibit 55 Global Freight Rates Have Declined in Real Terms by an Average of 0.9% p.a. Over the Last 260 Years...

Source: Stopford and Bernstein estimates and analysis.

21 Consequently, it is ironic that, of the "Big Three" global miners, Vale seems determined to undermine the pricing support of its key commodity and increase the risk of fragmenting global iron ore markets. As the most geographically distant from the bulk of iron ore demand (Asia), we estimate that China alone represented 56% of 2012 iron ore demand and produced just 13%, importing the balance of 42%. Consequently, Vale's profits and shareholders are the most vulnerable if the company's own cheap supply removes the need for high-cost inefficient artisanal mining in China. Vale's determination to bring huge low-cost supply onstream seems to indicate that its strategic direction is determined by incentives different than those of other publicly listed companies, perhaps due to the significant ownership by the Brazilian government (5.5% direct through Golden Shares and 34% indirect through the strategic consortium of Valepar).

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

1750

1760

1770

1780

1790

1800

1810

1820

1830

1840

1850

1860

1870

1880

1890

1900

1910

1920

1930

1940

1950

1960

1970

1980

1990

2000

2010

Re

al 2

000$

Global Freight Index 1750-2012

Access to Iron Ore Production That Is 2.5-5x More Efficient Than During the U.S. Industrialization Has Helped Enable the Rapidity of China’s Ongoing Development



AND LOVE THE ORE 47

Exhibit 56 …Leading to the Emergence of a Truly Global Market for Bulk Commodities, Including Iron Ore and Coal; Consequently, Relative Global Competitiveness in the Production of These Commodities Began to Matter

Source: Stopford and Bernstein estimates and analysis. Exhibit 57 Before the Advent of Scrap-Based EAF Production, U.S. Iron Ore Production Peaked

in the 30 Years Between 1950 and 1980

Source: USGS and Bernstein estimates and analysis.

0

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2,00019

00

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

Re

al 2

000$

Global Freight Index 1900-2012

0

20

40

60

80

100

120

140

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

Iro

n O

re —

Mt

US Iron Ore Production



AND LOVE THE ORE

Exhibit 58 The Transition to EAF-Based Production Saw Declines in Both Iron Ore Production and Employment

Source: BLS and Bernstein estimates and analysis. Exhibit 59 Consequently, the U.S. Iron Ore Productivity Peaked at Nearly 9,000 Tons/Man-Year

(Although for Much of the U.S. Industrialization, Its Intensity Was Far Lower)

Source: Govt. of Western Australia and Bernstein estimates and analysis.

In contrast, we show the development of iron ore mining in Western Australia

(see Exhibit 60 through Exhibit 62). Unlike the iron ore industry in the U.S., in Australia, it has never been intended for domestic consumption and has always had a strong international market orientation (Western Australian Government lifted its embargo on iron ore exports in 1960), focusing originally on post-WWII reconstruction and the subsequent industrialization of Japan and now on China. This development operated hand-in-hand with the declines in the international

0

5

10

15

20

25

30

35

40

45

5019

00

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

Th

ou

san

d P

eop

le E

mp

loye

d

US Iron Ore Mining Employment

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

To

ns

pe

r Man

-Yea

r

US Iron Ore Mining Productivity



AND LOVE THE ORE 49

freight costs, which enabled the economic movement of low-value material over significant distances. Consequently, the industrial development of the U.S. was predicated largely upon the productivity of its domestic mining industry. The U.S. did not and could not take advantage of the highest level of global raw materials production when it was industrializing and had to be content with the level of productivity that prevailed at the local or national level. So, while the U.S. iron ore intensity peaked at ~9,000 tons/man-year, Australia peaked at 20,000 tons/man-year. It is this level of productivity that the Chinese are also able to access (albeit indirectly) today, and the Japan and the Asian Tigers were able to enjoy during the later part of the 20th century.

Having said this, it is important to understand the difference between the relative productivities of Australia and the U.S. First of all, both American and Australian iron ore industries are very well capitalized, and it is capital rather than labor that stands behind raw material production (in contrast to China where mining is still labor driven). However, the two key differences are geology and population density. Different geological endowments: A miner in the U.S. needs to move about

twice as much run of mine (ROM) material as his Australian counterpart to generate the same amount of useful material. This is due to the difference between accessing direct shipping ore hematite (62% Fe grade) available throughout the Pilbara and magnetite (~35% Fe ROM) found in the U.S., which requires beneficiation and processing before it can be converted to pellets used by the U.S. steel makers.

Population density: Western Australia (ex-Perth) has, on average, 0.23 people per square kilometer versus 67 in Minnesota and 175 in Michigan. Virtually no one lives in Western Australia outside of Perth (see Exhibit 63). As a resource-intensive industry, mining requires significant land, water and power (not to mention there are those who have questioned the aesthetics of open-cut mining, as odd a suggestion as that may sound). A high population density always means that there will be competing calls on the resources that a miner needs as well as the potential for objections from local residents on aesthetic and environmental grounds. Consequently, it is much easier to mine in a sparsely populated area. On this note, China is 1,300 times more densely populated than the Pilbara!22

Before moving on to the labor productivity required to generate each ton of steel hitherto embedded in the Chinese capital stock, it is worth addressing why the declines in Pilbara mining productivity post 2006 are linked to the use of labor rather than declining asset productivity (which might initially appear to be the primary cause) — see Exhibit 62. It is easier to see what is going on if we look at the breakdown of production, employment and productivity by producer in the Pilbara (see Exhibit 64 through Exhibit 67). The large producers, Rio and BHP, have not suffered productivity declines anywhere near as severe as the average would suggest, though even these producers have suffered deterioration. However, for FMG, using the alternative surface mining technology, the productivity is about half that of the traditional mining methods. This indicates that the decline in productivity is largely attributable to the employment of labor on project and exploration sites for assets that are yet to produce rather than necessarily indicating a reduction in productivity from the mines already producing material. Consequently, Exhibit 68 shows the range of productivities that we use in the subsequent analysis.

22 Chinese statistic, excluding the provinces of Inner Mongolia, Xinjiang, Tibet, Qinghai and Gansu, as there is limited population and negligible mining.



AND LOVE THE ORE

Exhibit 60 In Contrast to the U.S., Western Australian Iron Ore Industry Was Always Driven by Exports — First to Japan and Today to China; the Declines in Shipping Costs Facilitated This Greatly

Source: Govt. of Western Australia and Bernstein estimates and analysis. Exhibit 61 Superior Geology and Continuing Improvement in Mining and Logistics Technology

Have Led to Substantial Increases in Mining Productivity


0

50

100

150

200

250

300

350

400

45019

60

1962

1964

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

Iro

n O

re —

Mt

Western Australia (Pilbara) Iron Ore Production

0

5

10

15

20

25

30

35

40

1960

1962

1964

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

Th

ou

san

d P

eop

le E

mp

loye

d

Western Australia (Pilbara) Iron Ore Mining Employment



AND LOVE THE ORE 51

Exhibit 62 In Addition, Low Population Density Renders Iron Ore Mining in Pilbara Hugely Productive...

Source: Govt. of Western Australia and Bernstein estimates and analysis. Exhibit 63 …And Provides a Significant Advantage Over the U.S. or China as Mining

Jurisdictions

Note: China excludes the "five provinces" of Inner Mongolia, Xinjiang, Tibet, Qinghai and Gansu.

Source: Wikipedia and Bernstein analysis.

0

5,000

10,000

15,000

20,000

25,000

1960

1962

1964

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

To

ns

pe

r Man

-Yea

rWestern Australia (Pilbara) Iron Ore Mining Productivity

295

175

67

0.89 0.23

China Michigan Minnesota Western Aus WA (ex Perth)

De

ns

ity

—P

ers

ons

pe

r Sq

uar

e km

Population Density by Iron Ore Mining Region



AND LOVE THE ORE

Exhibit 64 The Consolidation of the Industry Can Be Seen in Both the Concentration of Employment...

Source: Govt. of Western Australia and Bernstein estimates and analysis. Exhibit 65 ...And Production in Western Australia


14,278

9,712

5,974

1,240 1,177611 374 325 254 171 135 115 68 8

Peo

ple

Em

plo

yed

Western Australia Iron Ore Mining Employment (2011)

0

50

100

150

200

250

Iro

n O

re —

Mt

Western Australia Iron Ore Mining Production (2011)



AND LOVE THE ORE 53

Exhibit 66 Among the Majors, Rio and BHP Have Almost Identical Productivities and Are Almost Twice as Productive as Either FMG or Cliffs

Source: Govt. of Western Australia and Bernstein estimates and analysis. Exhibit 67 It Is Also Interesting to Note the Significant Declines in Productivity Seen at the

Company Level as Well as in the Industry Overall in Recent Years


26,901

16,397 16,011

8,7047,579 6,870 6,709 6,497

3,653 3,0772,035

To

ns

pe

r Man

-Yea

rWestern Australia Iron Ore Mining Productivity (2011)

-

20

40

60

80

100

120

140

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

100

= 20

02

Indexed Iron Ore Mining Labor Productivity

BHP Cliffs Resources Rio Tinto



AND LOVE THE ORE

Exhibit 68 However, the Decline in Productivity in Western Australia Is Due to the Inclusion of Labor That Is Allocated to Building Projects (With a Significant Production Lead Time) vs. Labor Allocated to Operations That Generate Current Volume


When turning to China and its recent rates of steel production and iron ore demand (see Exhibit 69 and Exhibit 70), it is important to stress how radically different its domestic iron ore production is from either the U.S. or Australia (see Exhibit 71 through Exhibit 75). Vast amounts of (hitherto) cheap labor historically stood behind the raw material production in China rather than the use of capital and machinery. However, post the late 1970s' reforms and the beginning of Chinese integration into the global economy (GATT and finally WTO membership), Chinese raw material consumption underwent a radical change. Imported raw materials rather than domestic production became the mainstay of industrial activity. The exports into China allowed the PRC to access the ever-increasing productivity of Australian iron ore miners (see Exhibit 76).

Since 1990, self-sufficiency in Chinese iron ore production has only been declining (see Exhibit 71). Given that the marginal producer of iron ore globally is in China (as discussed in the "If China Is to Grow Rich Before It Grows Old, It Must Finish Industrializing in the Next Decade" chapter), it is the high costs of domestic Chinese production that are currently setting raw material prices. Removing these producers will catalyze the end of the current secular period of pricing support (short-term supply demand fluctuations notwithstanding). In this regard, the trend decline is remarkably stable; if history is projected forward (admittedly, always a dangerous move), it would see China falling to 10% self-sufficiency (and thus becoming irrelevant to price formation) only by 2023.23 Clearly, extrapolating trend ignores the fact that conditions change and that supply and demand may accelerate, but due to the stabilizing effect of demand elasticity it may well be that the path to oversupply takes longer to come into play than many realize (see Exhibit 72).

However that may be, while the U.S. was reliant on its own geology and mining productivity to stand behind its industrialization, the decline in global

23 This assumes straight-line trend of declining Chinese self-sufficiency due to continued importation of iron ore into China by the majors at the same historical absolute amount of increased tonnage (rather than historical average growth rate of importation). It does not take into consideration the actual new project pipeline.

0

5,000

10,000

15,000

20,000

25,00019

60

1962

1964

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

To

ns

pe

r Man

-Yea

r

Western Australia (Pilbara) Iron Ore Mining Productivity

Productivity From Operations Productivity Including Projects

China Is Embedding Approximately One-Fourth the Labor Into Its Capital Stock Than the U.S. Did, But at a Steel Intensity That Is Approximately Double That of Peak U.S. Steel Consumption



AND LOVE THE ORE 55

freight costs enabled China to outsource this to Australia and take advantage of the superior geology and mining environment in the Pilbara. If we calculate the average productivity of the iron ore that China has actually consumed in its industrialization, we can see that it actually outstrips that of the U.S. (see Exhibit 77). The weighted average productivity24 of iron ore produced over the industrial history of the U.S. stands at 3,350 tons per man-year. The average productivity that stands behind the Chinese economic development is 8,750 tons per man-year. To return to Cheops, today we would be able to build 6,000 Great Pyramids, where the ancient Egyptians built one, employing exactly the same amount of manpower. This testifies to the increase in labor productivity over the intervening 5,000 years. Pyramid building in toto may be unproductive insofar as it satisfies no real human need, but that cannot be deduced from the relative intensity with which one society constructs pyramids versus another at a very different point in time.

Turning to Exhibit 78 and Exhibit 79, we see an incredibly strong relationship between relative productivity in raw material production and relative intensity of steel use between the U.S. and China. Chinese steel intensity is not much greater than that seen in the U.S. (or indeed anywhere else). Peak U.S. iron ore production occurred in 1951, at which time a U.S. miner could produce at a rate of 3,150 tons per year and steel intensity was 33kg/$'000. Today, steel intensity in China is 63kg/$'000 but the effective productivity of iron ore production is 13,600 tons/man-year. China is therefore embedding approximately one-fourth the labor into its capital stock that the U.S. did, despite having higher steel intensity in absolute terms. The more rapid embedding of steel in China's capital stock reflects not inefficiency per se, but rather that the productivity of the world's best iron ore miners (Australians) as opposed to just the geographically proximate producers can be utilized to advance the process of capital stock formation.

Exhibit 69 China Has Seen Both Its Steel… Exhibit 70 …And Iron Ore Production Increase

Source: WSA, NBS and Bernstein estimates and analysis. Source: WSA, NBS and Bernstein estimates and analysis.

24 Weighted by tons of iron ore.

0

100,000

200,000

300,000

400,000

500,000

600,000

700,000

800,000

Cru

de

Ste

el (k

t)

Chinese Crude Steel Production

0

200,000

400,000

600,000

800,000

1,000,000

1,200,000

Iro

n O

re (k

t)

Chinese Iron Ore Demand



AND LOVE THE ORE

Exhibit 71 The Iron Ore Majors (Vale, Rio and BHP) Have Aggressively Grown Their Volume Share of the Chinese Iron Ore Market Over the Last Decade, While Chinese Self-Sufficiency Has Declined Significantly

Source: WSA, NBS and Bernstein estimates and analysis. Exhibit 72 Extrapolating the Trend Decline in Self-Sufficiency Forward Suggests the End of the

Chinese Iron Ore Industry (and With It the High-Cost Price Support for Iron Ore) in the Region from 2020 to 2025; Clearly, Acceleration in Supply from the Seaborne Market and Declining Steel Growth Rates Would Advance This

Note: 2014-20 are Bernstein estimates.

Source: WSA, NBS and Bernstein estimates and analysis.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0

200,000

400,000

600,000

800,000

1,000,000

1,200,000

Ch

ine

se S

elf

-Suf

ficie

ncy

(%)

Iro

n O

re (k

t)

Chinese Iron Ore Supply

Imports Domestic Production Self-Suff iciency

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

2012

2014

2016

2018

2020

Se

lf-S

uff

icie

ncy

(%

)

Chinese Iron Ore Self-Sufficiency — Trend Decline



AND LOVE THE ORE 57

Exhibit 73 Chinese Growth in Iron Ore (on a Rich-Ore-Equivalent Basis) Peaked in 2007; Since Then, Declining Grades and Rising Costs Have Seen Useable Ore Production Decline

Source: NBS, WSA and Bernstein estimates and analysis. Exhibit 74 An Ever-Increasing Amount of Labor Has to Be Allocated to China's Iron Ore

Production…

Source: NBS, WSA and Bernstein estimates and analysis.

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

400,000

450,000

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

Iro

n O

re (R

OE

Bas

is) —

kt

Chinese Domestic Iron Ore Production

0

100

200

300

400

500

600

700

800

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

Th

ou

san

d P

eop

le E

mp

loye

d

Chinese Domestic Iron Ore Employment



AND LOVE THE ORE

Exhibit 75 …Leading to a Prolonged Decline in the Productivity of the Chinese Domestic Industry Since 2007

Source: NBS, WSA and Bernstein estimates and analysis. Exhibit 76 An Australian Miner Is 35x as Productive as His Chinese Counterpart; However,

Given That Globalization of the Iron Ore Industry Has Enabled China to Outsource Its Iron Ore Mining to Australia, Does That Differential Matter?


0

100

200

300

400

500

600

700

800

900

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

To

ns

pe

r Man

-Yea

r

Chinese Domestic Iron Ore Productivity

0

5,000

10,000

15,000

20,000

25,000

1960

1962

1964

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

To

ns

per

Man

-Yea

r

Western Australia (Pilbara) Iron Ore Mining Productivity



AND LOVE THE ORE 59

Exhibit 77 Productivity Uplift of the Iron Ore Imports Into China Has Enabled It to Overtake the U.S. in Terms of Effective Productivity in 2003

Source: WSA, NBS, BLS and Bernstein estimates and analysis. Exhibit 78 We Compare the Rate at Which Steel Is Being Embedded in China Relative to Peak

U.S. Steel Intensity (Y Axis) and the Peak Iron Ore Production (in Man-Hour Terms) to Which China Has Access Relative to That Accessible to the U.S. During Its Industrialization (X Axis); However, Due to Productivity Improvements, the Real Resources Consumed to Embed Steel in the Chinese Economy Are Markedly Lower Than Was the Case During the Industrialization of the U.S.


0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

To

ns

pe

r Man

-Yea

rEffective Chinese Productivity, Including Imports

USA Effective Chinese Productivity

R² = 0.8458

80%

90%

100%

110%

120%

130%

140%

150%

160%

20% 40% 60% 80% 100% 120% 140% 160% 180%Ch

ine

se S

teel

Inte

nsit

y to

Pe

ak U

S S

tee

l In

tens

ity

Ratio of Chinese Iron Ore Productivity to Peak US Iron Ore Productivity

Relative Iron Ore Productivity to Relative Steel Intensity



AND LOVE THE ORE

Exhibit 79 U.S. Iron Ore Production Peaked in 1951, With an American Miner Producing 3,150 Tons/Year and Steel Intensity of 33kg/$1,000; Today, China's Steel Intensity Is 63kg/$'000, While the Effective Productivity of Iron Ore Production Is 13,600 Tons/Man-Year


Imitation is often touted as the sincerest form of flattery, but we would argue that the outright purchase and adoption of another's technological developments go at least one step further. We mentioned earlier in this chapter that post the Open Door Policy and the beginnings of Chinese integration into the global economy (GATT and finally WTO membership), Chinese raw material consumption underwent a radical change, with imported raw materials rather than domestic production becoming the mainstay of industrial activity. This was not the only radical change that enabled greater steel embedding into the Chinese economy. The Chinese also purchased steel-making technology from the West; one of the more "famous" examples of this is the Famous Industrials Group, which began importing precision measuring tools and lock cylinders into China in the 1990s. Famous expanded into industry relocation projects; over the last 10 years, the company has relocated departments and production, including six complete industrial plants and production lines from the West into China.25 Some examples include: 1995 saw the purchase of the Bergkamen Coal Washing plant from the Monopol

mine (Northern Rhur Valley), its subsequent disassembly and transport to China. Today, it is run by SDIC Xinji Energy Co., Ltd., a Chinese coal producer with mines in the Anhui province (see Exhibit 80).

Three years later, in 1998, Famous purchased the Sophia Jacoba coal washing plant from the Hückelhoven coal mine in the Rhine Valley. Around 100 Chinese experts spent eight months disassembling the plant before relocating it to Eastern China, where it is also operated today by SDIX Xinji Energy Co (see Exhibit 81).

2002 saw the purchase of the Kaiserstuhl Coking Plant from the Dortmunder Westfalenhuette. The following year saw its disassembly by 300 Chinese specialists followed by its 2004 reassembly in Jinin, Shandong province. The

25 http://famous-germany.com/a_news.asp?classid=15.

0%

20%

40%

60%

80%

100%

120%

140%

160%

180%

50%

70%

90%

110%

130%

150%

170%19

95

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

Re

lati

ve Ir

on

Ore

Pro

duc

tivi

ty

Re

lati

ve S

teel

Inte

nsi

ty

Chinese Steel Intensity to Iron Ore Productivity

Chinese to Peak US Steel Intensity Chinese To Peak US Iron Ore Productivity

The Rate at Which China Has Been Embedding Steel Has Furthermore Been Accelerated by the Purchase of Western Steel-Making Technology



AND LOVE THE ORE 61

plant, which had only been operational for eight years prior to its purchase, recommenced operations in 2006 (see Exhibit 82).

It is worth noting that the purchase of entire plants and production lines is not limited to the mining and steel-making sector; Famous has also relocated spring factories (for the manufacture of trucks and railway vehicles), underground refrigeration plants, foot processing and packaging factories, and automation and communications systems, to name a few of its non-mining projects (see Exhibit 83). This process of disassembly and reassembly means that in a bare manner of months, not only did China benefit from the specifically relocated department or production itself, but the "specialists" involved in the dis- and re-assembly are now able to replicate and reproduce that technology on a wider scale across China. This further enables the rapidity of the Chinese industrialization process.

Exhibit 80 Bergkamen Coal Washing Plant Exhibit 81 Hückelhoven Coal Washing Plant

Source: Famous Industrial Group GmbH. Source: Famous Industrial Group GmbH.

Exhibit 82 Kaiserstuhl Coking Plant Exhibit 83 Werdohl Leaf Spring Factory

Source: Famous Industrial Group GmbH. Source: Famous Industrial Group GmbH.



AND LOVE THE ORE



AND LOVE THE ORE 63

Iron Ore Pricing Support Over the Next Decade Depends Upon Capital Discipline (Rio 360 Case Study)

On the supply side, as noted in the "Iron Ore 'Super-Cycle' Pricing Generated by High-Cost Producers Responding to the Rapidity of China's Industrialization" chapter, we see the inability of the Chinese mining sector to transition from labor-intensive to capital-intensive modes of production as a consequence of poor geological endowment supporting the current "super-cycle" pricing (this is, of course, bearing in mind that this failure is occurring against the backdrop of the rapid industrialization of China). As we see the costs on the supply side providing the long-term pricing structure, pricing support will persist, in our view, so long as a significant proportion of high-cost marginal producers in China (~35% of 2011 rich ore equivalent [ROE] Chinese supply, with marginal costs of production ~$120-180/t) are not displaced. Should the Western miners actively cease to underwrite continued Chinese industrialization (as the West underwrote its own and that of Japan), we see pricing support for iron ore continuing for the duration of China's industrialization (discussed in the "If China Is to Grow Rich Before It Grows Old, It Must Finish Industrializing in the Next Decade" chapter). If not, then it is 2015 that we see as the earliest time at which there will be sufficient new production coming onstream to displace high-cost producers.

We are aware that we are beginning to sound like a later day and mining-focused version of Cato the Elder (if such a thing can be imagined) and finish every piece with a cry that "capital is to be disciplined." Yet capital discipline is the single greatest source of value creation in the mining space — and this is particularly critical for iron ore, the fourth most crustally abundant element. For any large mining company, an "unrelenting focus" on value is synonymous with ensuring that capital is allocated to new volume only if it does not lead to price destruction. An “unrelenting focus” on shareholder value is diametrically opposed to any strategy designed (explicitly or not) to lower commodity prices.

This chapter continues with the "value maximizing growth strategies for large mining companies" theme. In particular, we focus on the possible 70Mtpa iron ore expansion by Rio Tinto as a case study and show how the decision to proceed (or not) with the project depends on how comfortable Rio Tinto owners are with the 3% steel demand growth target that the company's management clearly believes in. Shareholders in Rio Tinto may be comfortable risking an $18 billion loss on the chance of an $8 billion gain should China achieve 3% steel demand growth in a decelerating environment. Alternatively, they might wish to follow the example of the Emperor Augustus in considering the whole enterprise akin to fishing with a golden fish hook and that the gain, if all goes well, is too small to offset the risk of loss if it does not.

We further expand our analysis to demonstrate how globally superior value generation can result from the pursuit of locally optimal growth targets in consolidated markets, where industry incumbents are explicitly aware of the price destructive effects that excessive volume growth has. In itself, market consolidation and a "good" industry structure are not sufficient to drive superior returns. Instead, the miners require an approach to value that explicitly captures the price destructive effect of new volume. For us, the critical feature of the debate on Rio Tinto 360 is the emerging language from the company recognizing (for the first time) that the value of the investment proposition turns on the market impact that the extra tons

Capital Discipline Is the Single Greatest Source of Value Creation in the Mining Space; an "Unrelenting Focus" on Value Is Synonymous With Ensuring That Capital Is Allocated to New Volume Only If It Does Not Lead to Price Destruction



AND LOVE THE ORE

will have. In our view, it is a great leap forward and could precipitate a re-rating of the iron ore industry.

Capital Discipline Is Not the Same as Returns of Capital We have written previously and extensively on the need for capital discipline. In our call, Euro Metals & Mining: Not Enough Aristotle? The Fallacy of Growth & How Mining Companies Can Avoid Destroying Value, we showed how natural limits to growth emerge in consolidated commodity industries. Moreover, we believe that a natural evolution from "growth" to "value" exists in any mining company as it grows bigger. This transition corresponds to the development of a mining company from the entrepreneurial activity of new asset creation for a junior to the custodial function of value and asset preservation for a major. That is, the most important skill set for a mining junior is necessarily focused on the engineering and project-management aspects of mining. In contrast, once established as a major producer, it is the commercial, strategic and economic abilities that the miner's management will have tested.

Much has been made of capital discipline in the last few months, with CEO after CEO lining up to claim this virtue and to extol his "unrelenting focus" on shareholder value creation. However, what does capital discipline mean in practice, and why should it actually matter? After all, doesn't capital provide the lifeblood for continuing activity?

In theory, and ignoring taxation implications, shareholders should be ambivalent between receiving a dividend, which can be reinvested at their cost of capital, and having that dividend retained in the company to finance growth. If anything, it is always possible to generate a synthetic dividend stream,26 which — transaction costs aside — should render dividend policy a value-neutral proposition. Given this, capital discipline becomes a virtue only if capital allocation has gone wrong. Consequently, we view capital discipline in mining as a tool for preventing excessive organic growth. High commodity price preservation through maintaining a grip on supply is the single greatest value driver that mining companies have at their disposal. The greater the capital distribution returned to shareholders, the less likely a catastrophic oversupply is to result. In our view, mining will always be a cyclical business. However, it is up to the incumbent players to determine whether to feed or dampen that cyclicality. An "unrelenting focus" on value — as far as any large-scale miner is concerned — is synonymous with ensuring that capital is allocated to new volume only if it does not lead to price destruction.

For us, capital discipline is not about returns of capital per se. Instead, we see it as a mechanism to ensure that new volume growth is more accurately aligned with the forward-looking demand environment. We prefer our companies to run the risk of undersupplying the market and seeing higher prices, with the threat of possible demand destruction, to them oversupplying the market and guaranteeing prices will collapse. In other words, given the inherently asymmetric risk profile facing the miners, we would like them to play the odds in a manner that favors value creation for their own bottom lines and their shareholders. Consequently, we are not indifferent to the ways the capital is to be returned. Specifically, for us, capital discipline is not the same as asset disposal; we certainly would not like to see the miners execute wholesale asset disposals at (possibly) "fire sale" prices. Offloading assets to competitors is a risky strategic enterprise, which could result in de-consolidation of the market, while giving a potential competitor access to geology that it would not otherwise have (as we discuss later in this chapter, industry consolidation is one of the few means by which superior returns can be created in commodity markets). Moreover, in absence of synergies, shifting assets from one hand to another would create value only if the markets were inefficient

26 By either selling stock outright or writing out of the money call options on it to achieve the required income. From a risk-return perspective all of these alternatives are equivalent.

Capital Discipline Becomes a Virtue Only When Capital Allocation Has Gone Wrong; the Greater the Capital Distribution Out of the Mining Space, the Less Likely a Catastrophic Oversupply Is to Result



AND LOVE THE ORE 65

over a sustained period. In our view, such belief cannot become the strategy for the industry as a whole. Finally, asset disposals masquerading as capital discipline may enable the miners to sidestep the real objective, i.e., rationing of supply. Consequently, it is not returns of capital per se that we look for, but returns of capital that would otherwise be expended to accelerate supply into a declining demand environment. Clearly, we are not suggesting that an asset should be kept even if a truly attractive price is offered — rather, lucrative asset disposal opportunities should be explored all the time. Nevertheless, in our view, such exploration is more likely to yield attractive results at the top of the cycle rather than the bottom of it.

The mining industry is facing a dilemma at the moment. Although equity prices are reflecting a very significant and irreversible decline in commodity prices (most notably, for iron ore it sees the long-term price of US$80/t), the price environment, while volatile, is not yet utterly unsupportive of earnings. The last time equity valuations were as low as currently, the iron ore price had hit just US$85/t and it looked as if it were game over for that particular commodity. So why, when the 1H:13 iron ore price averaged US$129/t Aus FOB, is this not being reflected in equity valuations?

We believe that this comes back, again, to capital discipline. For retained earnings to be valued, they must have the prospect of being returned to shareholders at some future — even if far removed — date (hopefully, magnified many times over). Currently, however, the market fears that greater mining operating cash flows will translate to proportionally higher organic growth plans. The greater the capital spend today, the greater the volume tomorrow and — in a slowing demand environment — the greater the likelihood of a sustained and severe supply demand imbalance and price destruction. Consequently, higher earnings today imply lower earnings tomorrow. Furthermore, however counter-intuitive it may seem, the longer commodity price environment stays strong, and the more mining companies will de-rate unless they alter their behavior and convince the market that they are truly earnest about capital discipline. The problem is not the price environment, rather it is the strategy of the miners and their reaction to that price environment. Simply converting ever-expanding operating cash flows into new sources of production must come to an end. In our view, it will happen either in a value-accretive manner, with the miners willingly forgoing organic growth opportunities, or violently, when commodity prices are driven down to such a level that new organic growth is impossible, while the promise of a return on historical investment is proven illusory.

We strongly believe that increasing the dividend by 15% or 50% is simply insufficient. For the miners to see a structural re-rating, there must be a clear change in strategy that sees returns to capital rather than expenditure of capex as the priority. In 2012, Rio Tinto spent ~US$18 billion on capex, ~US$6 billion on corporate tax and ~US$3 billion on dividends. Why not simply reverse the priority of capex and dividends (see Exhibit 84 and Exhibit 85)?



AND LOVE THE ORE

Exhibit 84 Assuming Disciplined Capex, We Expect Rio Tinto to Have Room to De-Gear Its Balance Sheet Significantly Over the Next Few Years

Exhibit 85 Curtailment of Capex and De-Gearing of the Balance Sheet Would Enable Rio to Raise Its Dividend from US$3.0 Billion in 2012; We See No Reason Why It Should Not Double Over Time

Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis.

Rio Tinto Pilbara 360 Project — Does It Make Sense? In a meeting this past May that was closed to the media (and yet whose contents were known to the media almost immediately), Rio Tinto's management put forward a 70Mtpa Pilbara expansion to 360Mtpa to the company's board for approval in 4Q:13. We interpret this "leak" as an attempt by the company to gauge shareholder sentiment toward continued capital spend ahead of actually having to make the decision. While some pushback has resulted, rhetoric around value aside, we would expect management to execute this expansion if investor opposition grows muted. For us, the question is whether this expansion creates or destroys value for Rio Tinto and the mining sector.

In a recent interview, Rio's CEO Sam Walsh claimed that "the additional tonnage of 70 million tons was unlikely to result in a US$20/ton price decline."27 Critically, this statement represents the first time that the company explicitly recognized that the value of the investment may turn on the impact that it will have on the market. In our view, this recognition is a significant step forward for the industry, which moves the debate much closer to where it needs to be if returns from mining have any hope of being sustained into the medium term. While industry consolidation is a necessary requirement to see superior returns and genuine pricing power in a commodity industry, it is far from sufficient on its own. The necessary and sufficient condition is that there is industry consolidation as well as explicit awareness of the impact that new volume growth has on price. Pursuing industry consolidation — absent an understanding of how volume growth can act to undermine price and value — is merely a verisimilitudinous distraction. However, once this factor is understood, the prospect that superior returns can accrue naturally follows.

To make this point concrete, we look at the case study that Rio Tinto has provided in the form of its 360Mtpa Pilbara expansion. First of all, Rio Tinto, like

27 "Rio Said to Pursue $5 Billion Iron-Ore Project as Glut Looms," Bloomberg L.P., May 7, 2013.

2,775

45,182

38,577

18,861

4,071

8,807

19,41217,829

6,220

(3,729)

2006 A

2007 A

2008 A

2009 A

2010 A

2011 A

2012 A

2013 E

2014 E

2015 E

US

$m

Rio Tinto Net Debt

-

2%

4%

6%

8%

10%

12%

-

5,000

10,000

15,000

20,000

25,000

30,000

2012 A 2015 E

Div

iden

d Y

ield

(%)

US

$m

Rio Tinto Dividend Potential

Capex Tax Dividend Dividend Yield

Rio Tinto Owners Need to Get Comfortable With the Prospects for 3% Steel Demand Growth from China; It Is on This That the Value of the Expansion Turns; Will It Suppress Price Sufficiently to Render It Unattractive?



AND LOVE THE ORE 67

us, recognizes a significant inflection in the global cost curve (see Exhibit 86 and Exhibit 87). It should be noted that it is this inflection in a commodity cost curve that is responsible for both the margin that an industry generates as well as the volatility in that commodity's price. A steep cost curve implies significant industry profitability as well as significant risk; a flat cost curve has low profitability but a correspondingly low risk. As a result, the more valuable a commodity industry, the greater the requirement to ensure that volume growth is well managed and non-disruptive. Uninteresting and low-value businesses, on the other hand, can more or less look after themselves. Taking Rio Tinto's cost curve, we convert it from a discrete to a continuous mining output function for the industry as a whole, and from this curve, we calculate the elasticity of supply that is implicit in Rio Tinto's analysis. From this, we can see that Rio Tinto estimates the elasticity of supply to currently be somewhere between 3 and 4 (see Exhibit 88 and Exhibit 89).

Most importantly, an explicit awareness of price elasticity naturally generates the limits to organic growth. If one imagines an infinite reservoir of value-accretive projects, conventional wisdom would suggest that a company should execute as many of them as it can. Moreover, the same wisdom would suggest that there is a straight-line relationship between the number of projects that are executed (i.e., volume of output) and the value of that company (see Exhibit 90). However, this ignores that demand for a commodity is always finite and there is a negative relationship between commodity volumes and prices. The greater the volume in the market, the lower the price will be. Factoring that into the analysis of value maximization increasingly bends the straight line down into a parabola that shows a discrete point of maximum volume growth.

Exhibit 86 Rio Tinto's Cost Curve... Exhibit 87 ...Not Dissimilar to Our Own

Source: Rio Tinto. Source: AME, CRU, NBS and Bernstein estimates and analysis.

170

160

150

140

130

120

110

100

80

70

0

60

180

40

30

20

10

0

50

1,000,000750,000500,000250,000

90

1,250,000




AND LOVE THE ORE

Exhibit 88 Schematically, We Can Represent the Rio Tinto Cost Curve...

Source: Rio Tinto and Bernstein estimates and analysis. Exhibit 89 ...And So We Can Calculate the Elasticity of Supply That Rio Tinto Must Believe In

Source: Rio Tinto and Bernstein estimates and analysis.

0

20

40

60

80

100

120

140

160

180

200

0 200 400 600 800 1,000 1,200 1,400 1,600

Op

ex

(US

$/t)

Cumulative Iron Ore Production (mt)

Rio Tinto Iron Ore Supply

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

Ela

stic

ity


Rio Tinto Iron Ore Elasticity



AND LOVE THE ORE 69

Exhibit 90 The Value Destructive Effect of Excessive Volume Growth Can Be Seen Once the Impact of Supply Elasticity Is Considered

Source: Bernstein analysis and estimates.

Exhibit 91 summarizes this relationship. Given the elasticity that Rio Tinto

believes exists in the iron ore market, it ought to come to the conclusion that its expansion by 70Mtpa will lower the price of iron ore by ~US$27/t. Under such a scenario, it is simply expending US$5 billion of capital to reduce its total EBITDA by ~US$2.3 billion a year for a total value destruction of almost US$18 billion — not quite at Alcan levels, but getting there. This simple analysis can also be made more complex by building the full DCF model for the investment (or indeed by generating a closed-form algebraic solution to the problem as we did in Euro Metals & Mining: Not Enough Aristotle? The Fallacy of Growth & How Mining Companies Can Avoid Destroying Value) but the essence stays the same. One could object that the analysis uses a price of US$140/t and that this is far too high given that we "know" that price will fall to US$80/t. However, this begs rather than answers the question of whether or not to proceed with the investment and is actually irrelevant to the analysis in hand, which turns on price differences, not price levels.

0

20

40

60

80

100

120

140

0 25 50 75 100 125 150 175 200 225 250 275 300

Val

ue

(US

$bn

)

Iron Ore Volume Growth for New Entrant (Mt)

Value of Volume Growth for Different Price Elasticity

0.0 -0.5 -1.0 -1.5 -2.0 -2.5 -3.0

"Normal" straight‐line relationship between volume and value...but

only if elasticity is ignored



AND LOVE THE ORE

Exhibit 91 Rio Tinto's Claim That It Does Not See a Price Fall of US$20/t on the Back of Its 70Mtpa Increase in Production Becomes the Key Variable in Deciding Whether or Not the Project Makes Sense; If the Project Induces a Greater Decline, It Will Be Net Value Destructive for the Company


In any event, we also know that Rio Tinto believes in 3% iron ore demand growth from China. If we factor this in, the conclusion changes as shown in Exhibit 92. The extra demand growth is sufficient to limit the price declines to only ~US$14/t and so there is an incrementally positive contribution to EBITDA. That is, an expenditure of US$5 billion generates a positive value of ~US$8 billion. Clearly, everything here turns on whether the 3% demand growth in China is something that shareholders are happy with. Shareholders in Rio Tinto may be comfortable facing the loss of US$18 billion versus the gain of US$8 billion on the chance that China will achieve 3% steel demand growth in a decelerating environment. Alternatively, they might wish to follow the example of Emperor Augustus in considering the whole enterprise akin to fishing with a golden fish hook and that the gain, if all goes well, is too small to offset the risk of loss if it does not. Just as Augustus saw that there were limits to his own empire building and that, after all, Germany was just too small a prize to warrant the loss of Roman lives, so too we would caution that tons, once put in the market, are virtually impossible to remove. Much better to build only once there is clarity on the demand side from China. Festina lente, as Augustus would have it.

Original Volume Mt 290

Expanded Volume Mt 360

Change in Volume Mt 70

Current Chinese Iron Ore Demand Mt 1100

% Additional Supply from Rio % 6.4%

Price Elasticity - -3

Price Ex Expansion US$/t 140

Price Inc Expansion US$/t 113

Rio Tinto Unit Costs US$/t 35

Total Costs Ex Expansion US$bn 10.2

Total Costs Inc Expansion US$bn 12.6

Incremental EBITDA US$bn (2.3)

Multiple on EBITDA - 5.5

Less Capex Expended US$bn 5.0

Total Incremental Value US$bn (17.5)

Rio Tinto 360Mt Expansion Ex-Demand Growth



AND LOVE THE ORE 71

Exhibit 92 Rio Tinto Also Believes in a 3% Demand Growth World; This Would Be Sufficient to Limit the Price Impact of the Expansion to ~US$14/t and So Make It Appear Value Accretive (Though Clearly Not by the Degree That It Would Be If the Price Impact of the New Supply Were Neglected in Entirety)


If we were asked to provide a guide as to our preferred course of action with

this project, we would suggest taking the time from now until 4Q to really understand the true value of expansion to 360Mtpa under a range of demand scenarios and price responses. When it comes to the 4Q decision point, we would prefer Rio Tinto to push this back to 2Q:14, under the pretext of undertaking further market studies. This keeps the threat of new supply in the market, but increases the likelihood of continuing market tightness for a longer period without unduly upsetting its largest customer — China. Post this point, Rio Tinto can always come back with a revised path to growth, which shows a markedly slower ascent to the 360mt target, if demand is really there for it. Moreover, we believe that Sam Walsh has given himself an "out" on the project by declaring it contingent upon there being no "significant change in the demand-supply situation" (ironic in some sense, given that this project is in itself a significant change in the demand-supply situation). We would like to see this "out" utilized and, by doing so, see Rio Tinto's management demonstrate how superior industry structure can drive superior returns. Positive industry structure is not in itself a standalone positive outcome; it enables a positive outcome only in so far as it supports superior value creation. There is no point in trying to preserve market share if this leads to substantial value destruction.

Original Volume Mt 290

Expanded Volume Mt 360

Change in Volume Mt 70

Current Chinese Iron Ore Demand Mt 1100

% Additional Supply from Rio % 6.4%

% Chinese Demand Growth % 3.0%

Price Elasticity - -3

Price Ex Expansion US$/t 140

Price Inc Expansion US$/t 126

Rio Tinto Unit Costs US$/t 35

Total Costs Ex Expansion US$bn 10.2

Total Costs Inc Expansion US$bn 12.6

Incremental EBITDA US$bn 2.3

Multiple on EBITDA - 5.5

Less Capex Expended US$bn 5.0

Total Incremental Value US$bn 7.5

Rio Tinto 360Mt Expansion Including Demand Growth



AND LOVE THE ORE

The Virtue of Consolidation — Why Does It Matter? This simple example provided by Rio Tinto can be usefully extended to consider the case of commodity industries in general. In Exhibit 93 and Exhibit 94, we set out two industry models (deliberately chosen to be caricatures of the iron ore industry) that both show the potential for significant oversupply with possible new project growth in excess of demand growth. Both industries are currently generating significant margins, as is shown by the price elasticity. The only difference between these two industries is the degree of consolidation — one is highly consolidated (four industry players or the Herfindahl-Herschman Index [HHI] of 0.25, the threshold of "high concentration") and the other fragmented. Under each market situation, we simply ask what the locally optimal value maximizing strategy is for each incumbent player and then calculate the global response as the sum of the local responses. We then consider four possible scenarios: industry fragmentation versus consolidation and an explicit understanding or not of the value impact of the elasticity of supply.

In the first scenario (see Exhibit 95), the industry is fragmented where each participant acts on the belief that the only change that building a project has is to increase volume for the participant, rather than to lower price as well. Under this scenario, each player thinks that its volume will be value enhancing and so executes the maximum possible growth. Each believes that it will be able to add an incremental US$18 billion of value locally (or 27% versus the starting value of US$66 billion) — but the global effect of this local value maximization is massive oversupply and the destruction of US$32 billion of value for each player. It is the realization of this outcome that should see a de-rating of the industry from say 6x EBITDA to closer to 3x EBITDA as the market prices in today what it believes will come to pass tomorrow.

Exhibit 93 In Understanding the Value Impact of

Consolidation, We Construct Two Simple Industries — One Fragmented...

Exhibit 94 ...The Other Consolidated, But in All Other Respects Identical

Source: Bernstein analysis and estimates. Source: Bernstein analysis and estimates.

Industry Structure

Number of Players # 10

Size of Market Mt 1000

Market Share Mt 100

New Projects Available Mt 400

Max Growth per Player Mt 40

Demand Growth Mt 200

Price US$/t 150

Elasticity of Supply # -2

Caital Intensity for Growth US$/t 200

Opex US$/t 40

Industry Structure

Number of Players # 4

Size of Market Mt 1000

Market Share Mt 250

New Projects Available Mt 400

Max Growth per Player Mt 100

Demand Growth Mt 200

Price US$/t 150

Elasticity of Supply # -2

Caital Intensity for Growth US$/t 200

Opex US$/t 40

Industry Consolidation Is a Necessary But Not a Sufficient Condition for a Commodity Industry to Generate Superior Returns; the Virtue of Commercial Awareness Must Be Added



AND LOVE THE ORE 73

Exhibit 95 For Each Industry, We Simply Look at the Consequences of Each Player Looking to Maximize Apparent Local Value; in This Case, "Apparent" Means Simply Looking at the Economics of Their Own Decision Making; It Is Easy to Show That in a Situation of Potential Oversupply, Local Value Maximization Leads to Global Value Destruction for a Fragmented Industry With No Market Impact Awareness...


Under the second scenario (see Exhibit 96), the industry is still fragmented, but each participant has a more sophisticated understanding of value and recognizes that his individual project will impact price as well as volume. However, each project is simply too small to have much of an impact on price (in this case US$6/t) and so each player chooses to maximize volume growth just the same as before. While the prize that each player believes that it is aiming at is slightly lower than before (US$13 billion versus US$18 billion), it is still large enough to justify unconstrained volume growth. However, the net result is exactly the same as in the first example — massive oversupply, huge price cuts and value destruction, with the anticipation of such an outcome being presaged by an industry-wide de-rating.

Incumbent Player Total Industry

Player Ex Growth Total Industry Ex Growth

Volume Mt 100 Volume Mt 1000

Price US$/t 150 Price US$/t 150

Revenue US$bn 15 Revenue US$bn 150

Unit Costs US$/t 40 Unit Costs US$/t 40

Total Costs US$bn 4 Total Costs US$bn 40

EBITDA US$bn 11 EBITDA US$bn 110

EV/EBITDA Multiple - 6 EV/EBITDA Multiple - 6

EV US$bn 66 EV US$bn 660

Player Inc Growth Total Industry Inc Growth

Original Volume Mt 100 Original Volume Mt 1000

Growth Volume Mt 40 Growth Volume Mt 400








Less Capex US$bn 8 Less Capex US$bn 80

Total EV US$bn 84 Total EV US$bn 340

EV per Participant 34

Apparent Value Creation per Incumbent US$bn 18 Actual Value Creation per Incumbent US$bn -32

Fragmented Industry Excluding Awareness of Price Elasticity



AND LOVE THE ORE

Exhibit 96 ...And This Global Value Destruction Will Occur Whether or Not Each Individual Player Is Aware of the Impact That Its Decisions Have on the Market Price; the Apparent Gain from Organic Growth in Each Scenario Is an Overwhelmingly Compelling Prize


In the next scenario (see Exhibit 97), there is a consolidated market but each player is blithely unaware of the impact of its decisions on the market. Under this scenario, local value maximization is again realized through each player building in an unconstrained manner. It is clear that there is no difference in outcome between a fragmented and a consolidated industry, if the effect of price elasticity is neglected from the investment decision-making process (as seems to be the standard practice in mining). Industry consolidation is not an automatic good for shareholders — it is a positive only in so far as it yields a value outcome superior to that which would be achieved without consolidation. There is no point in trying to preserve an oligopoly if its preservation destroys rather than enhances value. In short, there is no such thing as a "good" industry structure if that structure does not return price power and superior value creation. And in mining, such an outcome cannot exist for as long as local value maximization is thought to be synonymous with ignoring the market impact of new volume. As so often in corporate life, the unanalyzed pursuit of "strategic" imperatives can easily lead to massive value destruction.

























Fragmented Industry Including Awareness of Price Elasticity



AND LOVE THE ORE 75

Exhibit 97 In a Consolidated Industry, Where Each Player Acts as If It Was Unaware of the Impact That Its Decisions Have on the Market, There Is Absolutely No Difference in Outcome Versus a Fragmented Industry; Exactly the Same Volume Is Brought to the Market and Exactly the Same Value Destruction Results; Consolidation in Commodity Markets Is Necessary But Not Sufficient to Generate Superior Returns


It is only when we come to the fourth and last scenario (see Exhibit 98) that a new feature emerges. Here again, local value maximization is sought; each agent individually builds as fast as it can commensurate with maximizing the value of its individual portfolio. However, because each player is independently aware of the fact that its decisions impact the market and because each has a large installed capital base, each comes to the conclusion that excessive growth will have a deleterious impact on value. Consequently, we see — for the first time — delivery of volume in a restrained manner. The effect of this at an industry-wide level is to generate value superior to that which each individual agent was targeting individually. Instead of massive oversupply, the "perfect" balance between new volume and price appreciation is met and industry-wide value is enhanced. Under this circumstance, the market should begin to re-rate rather than de-rate the sector, as awareness is formed about the emergent rationality of the industry and gets priced in.

We would like to stress no call is made here for concert or collusive action from the miners. Rather, a simple analysis of each agent's own local value maximization would be sufficient. In none of the foregoing analysis has anything other than local value maximization been the driving force behind decision making. There is no call to "look after" the market or to act to support price for other





Revenue US$bn 37.5 Revenue US$bn 150



EBITDA US$bn 27.5 EBITDA US$bn 110

















Consolidated Industry Excluding Awareness of Price Elasticity



AND LOVE THE ORE

participants. However, one has to look only at a number of media articles on the topic of Rio Tinto's 360Mt expansion target to realize how pervasive the misunderstanding of this topic is. The common error is the mistaken belief that looking to preserve price is somehow putting global rather than local value objectives at the heart of a company's strategy. This is not the case. Rather, putting price preservation at the heart of strategy is something that every company ought to do, but it is something that becomes of critical importance when the company approaches the scale of the large diversified miners.

Exhibit 98 The Sufficient Condition for Superior Returns in Consolidated Markets Is an Awareness of the Impact That Volume Decisions Have on Price and the Fact That This Changes the Value of the Entire Portfolio, Not Just the Marginal Ton; Under Such Conditions, the Pursuit of Local Optimality Returns a Global Value Solution in Excess of What Is Optimal


To summarize, industry consolidation is a necessary but not sufficient condition to enjoy superior returns in a commodity market (see Exhibit 99). Industry consolidation and an awareness of what local value maximization actually means are the necessary and sufficient conditions to enjoy pricing power and a superior return. The re-rating or de-rating that a sector should enjoy on the back of this is shown in Exhibit 100 through Exhibit 102. This shows how the rating varies by degree of market awareness and by degree of consolidation as well as certain





Revenue US$bn 37.5 Revenue US$bn 150



EBITDA US$bn 27.5 EBITDA US$bn 110
















Apparent Value Creation per Incumbent US$bn 24 Actual Value Creation per Incumbent US$bn 57

Consolidated Industry Including Awareness of Price Elasticity



AND LOVE THE ORE 77

other variables.28 It is clear that iron ore is an industry where there ought to be pricing power and superior return potential.29 However, it is clear (see Exhibit 103) that currently the miners are not getting the benefit of an expectation of superior performance — quite the reverse. Given that the industry is sufficiently consolidated to enjoy a premium rating (and will remain so under most scenarios), the reason for its absence must relate to expectations of how the market incumbents understand their role as agents in charge of value maximization. As we look at the sector, the critical flaw we see is that of incorrect or too simplistic valuation methodologies as we lay out in Euro Metals & Mining: Not Enough Aristotle? The Fallacy of Growth & How Mining Companies Can Avoid Destroying Value. The valuation policy error in mining companies results in poor capital discipline and an inability for globally superior outcomes to emerge from local value optimization. However, this is clearly beginning to change. Irrespective of the decision that is finally reached on Pilbara 360, the announcement from Rio Tinto's CEO Sam Walsh that is of key importance is that "the additional tonnage of 70 million tons was unlikely to result in a $20/t price decline."30 This is clearly an indication that the error in approach to value in at least one company is in the process of being rectified. It is for this reason that we continue to be optimistic about the future of the mining industry and continue to believe that emphasizing capital discipline is the key. It took, after all, 118 years to finally reduce Carthage; so perhaps a few years to see the emergence of real capital discipline is not too bad a result. We believe that a re-rating in this space is possibly precipitated by an appreciation that the defence of the iron ore price is a genuinely possible outcome. Exhibit 104 shows the impact of a re-rating around iron ore.

Exhibit 99 A Necessary Condition for Superior Price Performance Is That an Industry Be Consolidated, But It Is Not Sufficient; a Necessary and Sufficient Condition Is That There Is a Consolidated Industry Comprising Agents Who Understand the Price Impact of Their Own Decisions


28 The consolidation could have been illustrated using the Herfindahl-Herschman Index, but the effect is much the same as if we just use the number of market participants to illustrate the point. 29 While iron ore may be a scarce commodity, cape size export ports and world-class bulk transport infrastructure are not, and these are the key elements in the iron ore business. 30 "Rio Said to Pursue $5 Billion Iron-Ore Project as Glut Looms," Bloomberg L.P., May 7, 2013.

Conditions for Superior Returns in

Commodity Industries

Consolidated Fragmented

Aware of Price Impact of Own

Decisions Unaware of Price

Impact of Own Decisions



AND LOVE THE ORE

Exhibit 100 There Are Really Only Three Players That Matter in Iron Ore (Four at a Push)...This Is a Sufficient Enough Degree of Consolidation That the Pursuit of Local Value Maximization Ought to Embed Capital Discipline in the Industry and Generate Superior Returns

Source: Bernstein analysis and estimates. Exhibit 101 While the Degree of Possible Oversupply Is Clearly an Important Variable...


0

1

2

3

4

5

6

7

8

9

10

1 2 3 4 5 6 7 8 9 10

EV/E

BIT

DA

Mu

ltip

le o

f In

du

stry

Number of Players in Industry (i.e., Degree of Consolidation)

Multiple Uplift Available in Consolidated Commodity Markets Facing Oversupply

Including Market Aw areness Excluding Market Aw areness

0

1

2

3

4

5

6

7

8

9

10

1 2 3 4 5 6 7 8 9 10

EV/E

BIT

DA

Mu

ltip

le o

f In

du

stry


Multiple Uplift Available in Consolidated Commodity Markets Facing Oversupply

100% Possible Oversupply 75% Possible Oversupply 50% Possible Oversupply



AND LOVE THE ORE 79

Exhibit 102 ...Even More Important Is the Elasticity of Supply, and Here Rio Tinto Itself Estimates This to Be About Three; Moreover, It Is the Elasticity of Supply That Inflects Any Commodity Cost Curve and So Generates the Margins That an Industry Enjoys; the Higher the Elasticity, the Higher the Margin and Also the Lower the Degree of Consolidation Required to Generate Superior Returns

Source: Bernstein analysis and estimates. Exhibit 103 However, This Is Clearly Not What Is Being Discounted in the Stocks

Source: FactSet and Bernstein analysis.

0

2

4

6

8

10

12

1 2 3 4 5 6 7 8 9 10

EV/E

BIT

DA

Mu

ltip

le o

f In

du

stry


Multiple Uplift Available in Consolidated Commodity Markets by Elasticity

e=-1 e=-2 e=-3

4.65.0 5.0

5.8

Vale Rio Anglo BHP

EV

/EB

ITD

A

Current Trading Multiples for Iron Ore Producers



AND LOVE THE ORE

Exhibit 104 We Are Beginning to See Greater Clarity from the Miners in Their Thinking on Capital Discipline and Price Preservation...All of Which Ought to Create a Greater Chance of Seeing Growth Restrained and the Industry Re-Rate, Generating Significant Upside for the Miners, Especially Rio Tinto and Vale

Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis.

Rio BHP Vale Anglo

Current EV (US$m) 106,719 186,765 106,936 57,856

Proportion of Price Target Attributable to Iron Ore (%) 42.8% 25.4% 40.3% 24.6%

Current EV/EBITDA (x) 4.97 5.79 4.58 5.04

Bernstein Iron Ore Multiple (x) 6.18 6.81 7.41 4.87

Multiple Possible on Iron Ore (x) 7.68 8.31 8.91 6.37

Value Impact (US$m) 11,075 10,447 8,725 4,378

Share Price Impact ($) 6.0 2.0 1.7 3.4

Current* Share Price ($) 42.28 27.19 13.37 19.56

Per Share Impact (%) 14% 7% 13% 18%

* July 12, 2013



AND LOVE THE ORE 81

What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities?

The iron ore price (the most important source of mining EBITDA generation) has averaged US$135/t over the last five years and US$129/t average 1H:13 Aus FOB price — and yet mining company stocks have been among the worst performers of the year. This spring, we ran four fixed commodity price scenarios ("Grizzly," "Bear," "Bull" and "De Niro" in order of increasing aggressiveness) and examined how sensitive the value generated by our DCF (out to 2030) was to these different scenarios. The scenarios were bounded by the range of prices seen over the last five years (varying from the 25th percentile to the 80th percentile of the distribution), given the changes in the cost structure of the mining industry. We then selected the scenario that most closely resembled current market conditions and flexed the prices of the two most important commodities (from an EBITDA generation perspective for our coverage group) — copper and iron ore — in order to determine what the market was pricing in for these commodities. While the market still appeared comfortable about the longer-term copper prospects (baking into equity prices a 19% premium to the June 30, 2013 spot), the market was and remains deeply concerned about iron ore (baking in a 26% discount to the June 30, 2013 spot). Our analysis found that the market was pricing in iron ore at US$80/t and copper at US$8,000/t (see Exhibit 117 through Exhibit 121). We found — and continue to find — it difficult to believe that the pessimistic view on iron ore was tenable then or at the time of this Blackbook's publication.

We then asked ourselves "what Chinese growth outlook would be consistent with $80/t iron ore?" We calculated the Chinese GDP growth implied by consensus demand expectations in China, and then used this implied economic growth to back out the supply side response that, in concert with this demand expectation, would result in $80/t iron ore. We found that this was consistent with ~7% GDP growth — some 1.3% below the IMF trend forecast and 0.5% below the official government guidance — as well as significant new supply: ~550Mtpa of new iron ore capacity out to 2020 (a ~50% increase on the 1,123mt produced in 2012 globally), of which ~300Mtpa is surplus and hence price destructive.

Stress Testing DCF Valuations in Different Commodity Price Worlds At the time of this analysis in May, with the exception of the precious metals (most notably palladium), commodities were trading toward the bottom end of their historical five-year trading ranges (see Exhibit 105 through Exhibit 107), with the base metals and coal looking particularly depressed. We held all other variables constant (e.g., FX, capex and operating cost escalation) and focused on how our target price would vary with a changing long-term commodity price. Our "base case" factored in the most reasonable (in our view) temporal structure of the price line as a result of changing industry structures. Moreover, it is considered that, from a value perspective, the path taken to reach a long-term price is just as important as the overall price level.

Mining Equities Have Recently Been Baking in $80/t Iron Ore (~40% Below 1H:13 and Five-Year Average Prices) — This Is Consistent With Both ~7% Chinese GDP Growth (Below Both IMF Estimates and Chinese Targets) and ~550Mtpa New Iron Ore Capacity Out to 2020

Commodities Were Trading at Multi-Year Lows, So We Ran Four Commodity Price Scenarios to Examine the Sensitivity of Our DCF Valuation



AND LOVE THE ORE

Exhibit 105 Summary of Historical Commodity Prices

Source: Bloomberg L.P. and Bernstein estimates and analysis. Exhibit 106 Except for Some Precious Metals, Current Spot Prices of the Key Commodities for

the Miners in Our Coverage Were Substantially Below Their Most Recent Five-Year Mean

Source: Bloomberg L.P. and Bernstein analysis.

Nominal Commodity PricesPercentile of 5-Yr

Distribution

Nominal Commodity Prices 5Yr Avg YTD Spot YTD Spot

Platinum US$/oz 1,536 1,607 1,439 58% 28%

Palladium US$/oz 519 730 679 82% 73%

Gold US$/oz 1,282 1,609 1,414 76% 60%

Copper US$/t 7,312 7,814 6,975 58% 29%

Nickel US$/t 19,458 17,025 15,099 30% 13%

Zinc US$/t 1,971 1,995 1,813 48% 23%

Thermal Coal Ex Newcastle US$/t 102 91 86 34% 26%

Iron Ore Fines FOB Aus US$/t 135 146 138 58% 45%

Silver US$/oz 24 29 24 67% 52%

Uranium US$/lb 52 43 41 16% 3%

Molybdenum US$/lb 17 11 11 22% 21%

Aluminum US$/t 2,155 1,975 1,872 32% 20%

73%

60%

53%

46%

29% 29% 28%26%

21% 20%

13%

4%

Current Spot Price — Percentile of Five-Year Price Distribution



AND LOVE THE ORE 83

Exhibit 107 Higher Average Year-to-Date Prices Reflect Higher Commodity Prices in the Beginning of the First Quarter of the Year


In order to stress test this case and our stock recommendations, we took four price scenarios with fixed commodity prices (we label them "Grizzly," "Bear," "Bull" and "De Niro" in order of increasing aggressiveness) and examined how the price target generated by our DCF31 would vary with each of them.

A brief overview of our scenarios for a selection of the most important commodities is given in Exhibit 108 and Exhibit 109 and explained in detail in "Appendix 1: From Grizzly to DeNiro." In order to understand what these price scenarios represent, one should contrast them against the range of prices that we have seen over the last five years. Exhibit 110 and Exhibit 111 show the distributions of historical copper and iron ore prices (for the other major commodities, these are shown in Appendix 1). All the price scenarios we have chosen are bounded by the range of prices seen over the last five years (varying from the 25th percentile of the distribution, on average, for the "Grizzly" scenario to the 80th percentile for the "De Niro" scenario). Given the change in the cost structure of the mining industry over the last decade, if a price distribution further back than the last five years is taken, we believe it would lead to an overly negative assessment of what the future could hold. A copper price of US$6,500/t is markedly different in terms of impact now than it would have been in 2003.

31 Please note that we provide two valuation metrics: a one-year target price based upon our regression model and an intrinsic value generated by a DCF out to 2030. It is the latter that was used throughout this analysis. For more details on our valuation methodology, please see the "Valuation and Risks"' chapter.

82%76%

67%

58% 58% 58%

48%

34% 32% 30%

22%16%

Average YTD Price — Percentile of Five-Year Price Distribution

Our Four Scenarios Ranged from 25th to 80th Percentile of Historical Price Ranges



AND LOVE THE ORE

Exhibit 108 Summary of Commodity Price Scenarios vs. Historical Distribution — Real Terms

Source: Bloomberg L.P. and Bernstein estimates and analysis. Exhibit 109 Summary of Commodity Price Scenarios vs. Current Model — Real Terms


Real LT Commodity prices Percentile of 5Yr Distribution

2011 US$Current Model

Grizzly Bear Bull De NiroCurrent Model

Grizzly Bear Bull De Niro

Platinum US$/oz 1,486 1,500 1,700 1,900 2,100 33% 34% 72% 91% 98%

Palladium US$/oz 606 500 600 700 800 58% 46% 56% 77% 96%

Gold US$/oz 1,619 1,200 1,500 1,800 2,100 77% 45% 63% 98% 100%

Copper US$/t 7,652 6,500 7,500 8,500 9,500 51% 20% 44% 80% 94%

Nickel US$/t 15,991 16,000 18,000 20,000 22,000 17% 17% 40% 60% 71%

Zinc US$/t 1,841 1,800 2,000 2,200 2,400 26% 21% 48% 71% 90%

Thermal Coal ex Newcastle US$/t 88 80 90 100 110 29% 18% 33% 56% 64%

Iron Ore Fines FOB Aus US$/t 74 70 90 110 130 11% 7% 19% 26% 36%

Silver US$/oz 30 20 25 30 35 69% 50% 53% 69% 90%

Uranium US$/lb 44 45 50 55 60 24% 27% 45% 73% 81%

Molybdenum US$/lb 11 10 12 14 16 15% 7% 28% 42% 64%

Aluminum US$/t 2,405 1,700 1,900 2,100 2,300 74% 11% 22% 49% 67%

Real LT Commodity Prices Scenario as % of Current LT Price

Current Model

Grizzly Bear Bull De Niro Grizzly Bear Bull De Niro

Platinum US$/oz 1,486 1,500 1,700 1,900 2,100 101% 114% 128% 141%

Palladium US$/oz 606 500 600 700 800 83% 99% 116% 132%

Gold US$/oz 1,619 1,200 1,500 1,800 2,100 74% 93% 111% 130%

Copper US$/t 7,652 6,500 7,500 8,500 9,500 85% 98% 111% 124%

Nickel US$/t 15,991 16,000 18,000 20,000 22,000 100% 113% 125% 138%

Zinc US$/t 1,841 1,800 2,000 2,200 2,400 98% 109% 119% 130%

Thermal Coal ex Newcastle US$/t 88 80 90 100 110 91% 103% 114% 126%

Iron Ore Fines FOB Aus US$/t 74 70 90 110 130 94% 121% 148% 175%

Silver US$/oz 30 20 25 30 35 67% 83% 100% 117%

Uranium US$/lb 44 45 50 55 60 101% 113% 124% 135%

Molybdenum US$/lb 11 10 12 14 16 91% 109% 127% 145%

Aluminum US$/t 2,405 1,700 1,900 2,100 2,300 71% 79% 87% 96%



AND LOVE THE ORE 85

Exhibit 110 The "Fat Tail" in the Copper Price Distribution Relates to the Period Immediately After the Financial Crisis; We Would Interpret the Shape of the Price Distribution as Telling Us Something About the True Costs of Marginal Production

Source: Bloomberg L.P. and Bernstein analysis. Exhibit 111 The Shift in Pricing Regime from the Benchmark to the Spot Market as Well as the

Deterioration in Chinese Domestic Mined Iron Ore Grades Result in a Markedly Different Price Distribution for Iron Ore Than for Copper


Please refer to Appendix 1 of this Blackbook for other commodity price distributions.

0%

2%

4%

6%

8%

10%

12%2,

000

2,25

0

2,50

0

2,75

0

3,00

0

3,25

0

3,50

0

3,75

0

4,00

0

4,25

0

4,50

0

4,75

0

5,00

0

5,25

0

5,50

0

5,75

0

6,00

0

6,25

0

6,50

0

6,75

0

7,00

0

7,25

0

7,50

0

7,75

0

8,00

0

8,25

0

8,50

0

8,75

0

9,00

0

9,25

0

9,50

0

9,75

0

10,0

00

Pe

rce

ntil

e

Cu Spot Price — US$/t

Five-Year Nominal Copper Price Distribution

Spot as of 22/04/2013 $6,975/t

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200

Pe

rce

ntil

e

Fe Spot Price — US$/t

Five-Year Nominal Iron Ore Price Distribution

Spot as of 22/04/2013 $138/t



AND LOVE THE ORE

Exhibit 112 gives the high-level summary of how each of our stocks would respond to the valuation scenarios used. Vale is tied with Rio Tinto for the lowest downside in our Grizzly scenario and shows the highest upside in Bull and De Niro scenarios; the company with the greatest downside across all scenarios is Glencore. This was largely a consequence of our then-negative view on the long-term iron ore price32 (albeit with an interesting trajectory down to this level) versus copper and the high level of operational gearing in Vale on the back of its significant capex program (see Exhibit 113 through Exhibit 116).

For each of our coverage companies, we detail the pro forma financials for each of the scenarios, giving a summary income statement, balance sheet and cash flow. The full details of these scenarios are given in Appendix 1, but for the sake of brevity only the headline details have been included here. We also show how the SOTP DCF value varies by scenario for our coverage, as we base our target prices on EV of the aggregate assets held by each company.

Exhibit 112 Vale Shows the Greatest Potential Upside and Widest Price Distribution Under Our

Scenarios (on the Back of Operational Gearing), While Glencore Shows the Greatest Downside and Least Upside Relative to Our DCF Value (as a Result of Our Bullish View on Copper in the Medium Term)


Exhibit 113 Rio Tinto Shows the Least Downside to Our Scenarios and Glencore the Greatest

Exhibit 114 As Soon as a More Positive Iron Ore Price Is Introduced, the Impact of Operational Gearing in Vale Is Immediately Apparent


32 Please see the "Our Iron Ore Price Forecast" chapter for our updated iron ore price forecast.

TP / NPV per Share vs. Current Model

Current Model


Anglo American £ 23.00 6.51 17.03 26.84 36.22 -72% -26% 17% 57%

BHP Billiton £ 25.00 6.54 16.71 27.02 37.49 -74% -33% 8% 50%

Glen Xta Pro Forma £ 5.16 1.21 3.05 4.90 6.75 -77% -41% -5% 31%

Rio Tinto £ 45.75 15.23 37.73 59.99 81.69 -67% -18% 31% 79%

Vale BRL 29.25 9.73 34.38 58.97 83.54 -67% 18% 102% 186%

-67% -67%

-72%

-74%

-77%

Rio Tinto Vale Anglo American

BHP Billiton

Glen Xta Pro Forma

TP vs. Current Model — Grizzly

18%

-18%

-26%

-33%

-41%

Vale Rio Tinto Anglo American

BHP Billiton

Glen Xta Pro Forma

TP vs. Current Model — Bear

Vale Showed the Greatest Upside Across All Scenarios; Glencore the Lowest



AND LOVE THE ORE 87

Exhibit 115 Our Target Prices Are the Equivalent of the "Bull" Scenario Even Though We Reach These Targets With Higher Short-Term Prices and Lower Long-Term Prices

Exhibit 116 It Is Difficult for Us to See Further Upside in Our Target Price for Glencore as It Already Factors in a Very Strong Copper Price



What Iron Ore Price Is Being Baked In? In addition to stress testing our equity valuations, we also aimed to find out what iron ore and copper prices the market was pricing into the miner's equity prices. From the four scenarios run earlier, we selected the scenario that most closely resembled current market conditions ("Bear") and flexed the prices of the two most important commodities (from an EBITDA generation perspective for our coverage group) — copper and iron ore — in order to determine what the market was pricing in for those commodities. We found that mining equity valuations were baking in $80/t iron ore, a ~40% discount to year-to-date and five-year average prices (see Exhibit 117 through Exhibit 121).

Exhibit 117 As This Scenario Analysis Run in Late May Shows, Rio Tinto Offers the Most

Attractive Balance Between Risk and Potential Upside

Note: Price current as of May 21, 2013, when this analysis was run.


102%

31%

17%8%

-5%


BHP Billiton

Glen Xta Pro Forma

TP vs. Current Model — Bull

186%

79%

57%50%

31%


BHP Billiton

Glen Xta Pro Forma

TP vs. Current Model — De Niro

Rio Tinto (GBP), Current Price = £28.78 Iron Ore (US$/t)

TP 70 90 110 130

6,500 20.76 35.94 50.73 65.28

7,500 22.57 37.73 52.50 67.03

8,500 24.38 39.52 54.28 68.79

9,500 26.21 41.33 56.06 70.55Co

pp

er

(US

$/t)

Flexing Our Price Assumptions Revealed That Iron Ore Was Priced in at a ~40% Discount Year-to-Date and Versus the Five-Year Average



AND LOVE THE ORE

Exhibit 118 The Presence of Oil in the BHP Portfolio Adds a Third Critical Variable That Renders It Harder to Unpack the Implied Iron Ore and Copper Price from the Market Price; Oil Remains a Key Diversifier for the Highest Quality Stock in Our Coverage


Source: Corporate reports and Bernstein estimates and analysis. Exhibit 119 While the Market Price of Vale Might Seem to Imply a Higher Iron Ore Price, We

Believe That This Is a Function of Us Modeling a Higher Tax Rate for Vale Than the Market


Source: Corporate reports and Bernstein estimates and analysis. Exhibit 120 For Anglo, the Issues of Both Platinum and South African Political Risks Add a Layer

of Complexity to the Investment Proposition



BHP Billiton (GBP), Current Price = £19.12 Iron Ore (US$/t)

TP 70 90 110 130

6,500 11.22 15.36 19.50 23.66

7,500 12.57 16.71 20.87 25.03

8,500 13.93 18.07 22.23 26.40

9,500 15.29 19.44 23.60 27.77Co

pp

er

(US

$/t)

Vale (BRL), Current Price = BRL32.1 Iron Ore (US$/t)

TP 70 90 110 130

6,500 13.14 33.15 53.15 73.14

7,500 14.37 34.38 54.38 74.37

8,500 15.40 35.41 55.41 75.41

9,500 16.39 36.41 56.41 76.40Co

pp

er

(US

$/t

)

Anglo (GBP), Current Price = £15.49 Iron Ore (US$/t)

TP 70 90 110 130

6,500 12.32 15.05 17.75 20.43

7,500 14.33 17.03 19.69 22.34

8,500 16.33 18.99 21.62 24.24

9,500 18.31 20.93 23.54 26.13Co

pp

er

(US

$/t

)



AND LOVE THE ORE 89

Exhibit 121 Note That for Glencore Xstrata, We Run the Analysis on Our Combined Pro Forma Glencore Xstrata; Copper Is the Single Greatest Sensitivity for Glencore Xstrata; to See Outperformance in This "Not Yet Institutional Quality" (from a Governance and Reporting Perspective) Stock, It Is Necessary to Believe in a Strong Copper Price



What GDP Outlook Is Consistent With $80/t Iron Ore? Given that the equity market appeared to be discounting a US$80/t iron ore world, we then asked ourselves what the world needs to look like for these prices to come about. Specifically, "what growth outlook for the world's largest commodity consumer, China, would be consistent with $80/t iron ore?"

We calculated the Chinese GDP growth implied by consensus demand expectations in China, and then used this implied economic growth to back out the supply side response that, in concert with this demand expectation, would result in $80/t iron ore.

We found that US$80/t iron ore is consistent with a ~7% GDP growth — some 1.3% below the IMF trend forecast and 0.5% below the official government guidance. Our framework demonstrates that consensus expectations therefore imply ~550Mtpa of new iron ore capacity out to 2020, of which ~300Mtpa comes in excess of the requirement and is price destructive. Yet more evidence of the pressing need for a more restrained approach to organic growth by the majors!

Because China is the largest consumer of iron ore globally, and is only one-third of the way through the most rapid industrialization ever seen globally, the most important demand-side question for mining stocks is "How does one estimate future steel intensity in China?" Exhibit 122 shows the history of Chinese steel intensity as well as a plausible forward-looking trend line. The problem, however, is that a lot of such lines might look equally convincing and yet each would represent a substantially different future for the Chinese steel. There must be more to determining the future of this key variable than the aesthetics of different curves, and yet so often this still is the basis of demand projections.

Simply examining steel intensity curves does not provide enough information to decide if China is "on track" relative to other industrialized countries. Instead, we require a means to link steel intensity today with steel intensity tomorrow. We find such a link in the constant relationship between an economy's steel capital stock and its overall economic activity (R-squared = 65%) — see Exhibit 122 and Exhibit 123. Although steel intensity has varied from industrialization to industrialization (the U.S. was twice as intensive as France, and China is again some 50% more intensive than the U.S.), it has always been the embedding of steel in physical infrastructure that enabled a transition from an agricultural economy to one dominated by services. We therefore use this relationship to forecast how China's capital stock is likely to evolve.

In order to ground our analysis, we take the trajectory of U.S. steel capital stock formation relative to output as a benchmark development path for any

Glen Xta (GBP), Current Glencore Price = £3.37 Iron Ore (US$/t)

Implied Glen-Xta NPV 70 90 110 130

6,500 2.49 2.49 2.49 2.49

7,500 3.05 3.05 3.05 3.05

8,500 3.61 3.61 3.61 3.61

9,500 4.17 4.17 4.17 4.17Co

pp

er

(US

$/t)

What Is the Implicit Chinese GDP Assumption Consistent With US$80/t Iron Ore? ~7% — Some 1.3% Below the IMF Forecast and 0.5% Below the Official Government Guidance

U.S. Industrialization Provides a Canonical Pattern from Which to Derive the Multiplier Between Economic Growth and Metal Demand Growth for China



AND LOVE THE ORE

economy (see Exhibit 124 and Exhibit 125). These exhibits show the increasing levels of productivity in the U.S. — that is, incrementally higher levels of output being generated from each additional unit of input. In fact, the relationship between steel and output measures the evolution of productivity within an economy. However, increases in productivity have only become evident after the U.S. reached a steel stock of 10,000kg/capita versus ~4,500kg/capita in China today. Although China remains more steel intensive than the U.S., efforts by the Chinese government to diversify its economy away from reliance upon cheap outsourced manufacturing result in a trajectory that is directionally converging with that of the U.S. (see Exhibit 126).

In our annual forecasts, we extend the pattern of convergence seen historically with ever-increasing productivity in the Chinese economy. While a case could be made that China should be more metal intensive than the U.S. (due to its population density and intensity of housing and the associated infrastructure), we believe that the U.S. represents a sensible base case. Moreover, a more metal-intensive China would only defer rather than obviate the need for the transition that was experienced by the U.S. In any event, this adds only upside risk to our steel forecast.

Exhibit 122 A Central Question for an Iron Ore Price Forecast Is How to Project China's Steel Intensity Into the Future? Looking at Historical Steel Intensity Alone Is Insufficient

Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis.

0

10

20

30

40

50

60

70

80

0 5 10 15 20 25 30 35 40 45 50

Ste

el In

ten

sity

(kg

/$'0

00

GD

P)






AND LOVE THE ORE 91

Exhibit 123 The Relationship Between Economies' Capital Stock and Level of Output (R-squared = 65%) Provides the Missing Data Point for Forecasting How the Capital Stock to Output Ratio Will Evolve in China


R² = 0.6532

0

10

20

30

40

50

60

70

80

0 5,000 10,000 15,000 20,000

Ou

tpu

t —

GD

P/C

apit

a $'

000

(Rea

l 20

05 P

PP

$)

Capital Stock — kg Steel/Capita (including depreciation)

Capital Stock to Output by Country

China 2012

UK

Kuwait

Germany

Taiwan

USA

China 2020



AND LOVE THE ORE

Exhibit 124 We Use the U.S. Development Path as a Basis for China's Industrialization Forecast

Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis. Exhibit 125 The Benefit of Increasing Productivity Over Time Is Easier to See If the Axes in the

Earlier Graph Are Inverted; in the U.S., the Real Advances Took Place Only Once the Capital Stock Had Been Built and Urbanization With the Accompanying Labor Force Transformation Had Been Completed


0

5

10

15

20

25

30

35

40

45

50

- 2,000 4,000 6,000 8,000 10,000 12,000 14,000

US

GD

P/C

ap

ita

$'0

00

(Re

al 2

005

PP

P)

Kg of Steel in Economy per Capita

US Steel Capital Stock

R² = 0.9595

-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

0 5 10 15 20 25 30 35 40 45 50

Kg

of

Ste

el in

Eco

no

my

pe

r C

apit

a

US GDP/Capita $'000 (Real 2005 PPP)

US Steel Capital Stock (Inverted Axes)



AND LOVE THE ORE 93

Exhibit 126 China Is More Steel Intensive Than the U.S., But Shows Convergence


We use this convergence with the U.S. levels of steel productivity to calculate

a consistent steel intensity for any level of overall economic activity. For illustration only, we show a steel-intensity path that would enable China to reach roughly current levels of the U.S. output by 2020. It would require China to post GDP growth rates of ~20% for the next seven years, and, unsurprisingly, such a growth pattern would show a very rapid convergence to the U.S. (see Exhibit 127). In order to support this level of economic activity and to transition its large rural population into urban centers, China needs to embed steel. The total amount of new steel that would be required is supplied by the relationship between output and capital stock. Assuming this amount of new steel would have to be generated over the next seven years, we can then calculate the run rate of steel production (and hence steel intensity) for the economy as a whole. This, thus, links current and future steel intensities. Clearly, under the implausible/illustrative scenario of 20% GDP growth, there must be a correspondingly large increase in steel intensity (see Exhibit 128).

A further sensitivity should be flagged. The database of industrial data we have assembled for all countries currently reporting to the WSA, dating back to 1900, enables us to choose any historical paradigm of steel productivity as the basis for Chinese development. Consequently, we could show China converging to Japanese or German levels of steel productivity.

However, given the importance of the U.S. economy to global demand and global trade, we believe that the development patterns of countries other than the U.S. are less representative for China — the soon-to-be largest economy in the world.

-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

0 5 10 15 20 25 30 35 40 45 50

Kg

of

Ste

el p

er C

apit

a

GDP/Capita (Real 2005 PPP)

Steel Capital Stock — China vs. USA

USA China



AND LOVE THE ORE

Exhibit 127 For Illustration Only, We Show a Development Path That Would Enable China to Converge to the Current U.S. GDP Levels by 2020; It Would Require a 20% Growth Rate Each Year

Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis. Exhibit 128 Using Convergence as the "Missing" Data Enables Us to Calculate the Transition

Path Between Steel Intensities at Two Different Points in Time and Derive the Relationship Between Overall Economic and Metal Growth Rates


-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

0 5 10 15 20 25 30 35 40 45 50

Kg

of S

teel

per

Cap

ita


Steel Capital Stock — China Forecast vs. USA

USA China Actual China Forecast

0

10

20

30

40

50

60

70

80

- 5 10 15 20 25 30 35 40 45 50 55

Ste

el In

ten

sity

—K

g/$

'000

Output — GDP per Capita $'000 (Real 2005 PPP)

Steel Intensity Forecast

China Actual China Forecast Original SCB Trend



AND LOVE THE ORE 95

At our September 2012 launch, we used the April 2012 IMF global GDP forecasts as the basis for our demand calculations. Since our launch, the Chinese economy has unquestionably slowed down. In April 2012, the IMF was forecasting an average annual growth of 8.7% over 2014-2016. This estimate was revised down to an average 8.5% over 2014-2016 as of 2Q:13 (see Exhibit 129 and Exhibit 130). Furthermore, at the time of our launch, we were happy to calibrate China's future steel intensity forecast looking at the transition that would be required out to 2020 and beyond (as shown in Exhibit 123). However, in the face of frequent growth revisions, looking at only longer-term trends is insufficient. Instead, an explicit year-by-year analysis and greater rigor is required to explicitly link economic activity to steel use.

Taking the April 2013 IMF GDP deck as a starting point, the path to convergence for the U.S. is shown in Exhibit 131. The resultant steel intensity is then shown in Exhibit 132. Evidently, steel intensity shows a marked departure from the too simple trend line we had drawn previously (yet again, this highlights the dangers, if any further highlighting was needed, of too simple curve plotting in commodity markets analysis). However, if the acceleration of the Chinese economy back up to a trend rate of 8.5% is believed — and this is the IMF's current base case for 2016-2019 — steel production will accelerate accordingly (see Exhibit 132). Given the relative immaturity of the Chinese economy, it is still at a very metal-intensive phase of its development. Consequently, accelerating growth requires acceleration in raw material consumption. The rising productivity of the U.S. economy, which enabled it to grow output with no corresponding increase in metal intensity, occurred only after its capital stock had been built. China is still some way from this level (~4,500kg of steel per capital in China versus 12,000kg in the U.S.). Consequently, in order to accelerate growth without accelerating metal intensity, China's economy would have to experience an unprecedented increase in metal productivity. In any event, redrawing the trend steel intensity line that incorporates these more rigorously derived data points results in the structure shown in Exhibit 133. Having derived the steel intensity for the Chinese economy, it is possible to calculate the total steel production and volume growth (see Exhibit 134). More importantly, the ratio between metal growth and economic growth can be calculated from the first principles (see Exhibit 135). Instead of asserting the relationship between metal growth and economic growth, which simply masks rather than solves the problem, this approach enables it to be derived consistently.

The IMF's GDP Deck Implies That China's Steel Intensity Would Need to Accelerate Again



AND LOVE THE ORE

Exhibit 129 Since the Time of Our Original Analysis, China's Economy Has Slowed Down Markedly

Exhibit 130 In the Space of a Year, the IMF Revised Its Growth Projections Down by 0.5% on Average Over 2014-16)

Source: IMF and Bernstein analysis. Source: IMF and Bernstein analysis. Exhibit 131 Taking the IMF's April 2013 Growth Forecast, We Estimate the Following

Convergence to the U.S. Steel Stock Levels for China…


6%

7%

8%

9%

10%

11%

12%

13%

14%

15%

IMF China GDP Forecast

Apr-12 Apr-13

-0.8%

-0.7%

-0.6%

-0.5%

-0.4%

-0.3%

-0.2%

-0.1%

0.0%

0.1%

IMF China GDP Changes

-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

- 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0

Kg

of

Ste

el p

er

Ca

pit

a


Capital Stock of Steel — China Forecast vs. USA




AND LOVE THE ORE 97

Exhibit 132 …Which Would Imply the Following Trajectory for Steel Intensity; the Previous Steel Intensity Trend Line Was Extrapolated in Order to Keep the Long-Term Evolution of China Consistent With the U.S.; Now, We Calculate It Year-by-Year

Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis. Exhibit 133 We Refine Our Previous Steel Intensity Trend Line…


0

10

20

30

40

50

60

70

80

- 5 10 15 20 25 30 35 40 45 50 55

Ste

el In

ten

sit

y —

Kg

/$'0

00




0

10

20

30

40

50

60

70

80

- 5 10 15 20 25 30 35 40 45 50 55

Ste

el I

nte

ns

ity

—K

g/$

'000



China Actual China Forecast (IMF GDP) Original SCB Trend Rev ised IMF/SCB Trend



AND LOVE THE ORE

Exhibit 134 …And Generate the Following Steel Production Forecast; It Shows Declines in Crude Steel Growth Seen Post 2020; in Our View, It Is Unavoidable Under All Chinese Growth Scenarios

Source: WSA, IMF, Mitchell, Maddison and Bernstein estimates and analysis. Exhibit 135 Instead of Having to Guess the Relationship Between Metal Growth and Overall

Economic Growth, We Can Derive It from the First Principles


-5.0%

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

0

200

400

600

800

1000

1200


Cru

de

Ste

el P

rod

uct

ion

Gro

wth

Cru

de

Ste

el —

Mt

Implied IMF Steel Production and Growth

-0.50

0.00

0.50

1.00

1.50

2.00

2.50

200

5A

200

6A

200

7A

200

8A

200

9A

201

0A

201

1A

201

2A

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

Chinese Steel Growth to GDP Growth Multiplier — IMF Scenario



AND LOVE THE ORE 99

Inverting the earlier calculation, we can unpack the assumptions consistent with a particular view on China's expected GDP and, by extension, steel demand growth. We start with consensus expectations for steel production and price. We take the steel forecast of CRU — a widely used high-quality data provider to the mining industry (see Exhibit 136). It is one of the most frequently used forecasts in the industry.

We derive the GDP deck consistent with the steel demand evolution assumed by CRU (see Exhibit 137). The Chinese GDP growth rate that is consistent with consensus expectations for steel demand is therefore ~7% — some 1.3% below the IMF trend figure and 0.5% less than the official government target. Historically, the target has represented a floor to the achieved growth rather than an average expectation and is a figure that is not subject to frequent revisions, which might cause loss of face within the government. Moreover, the target aims to balance the competing objectives of avoiding inflation, while ensuring employment (clearly, something easier in theory than practice). Consequently, we believe that consensus expectations embed a bear case, rather than an equally weighted upside to downside scenario.

Nevertheless, the consequences that such steel demand path would have are evident in Exhibit 138 and Exhibit 139. The consensus expectation sees the decline in steel intensity seen over the last four years continuing into the future. We would like to stress that Chinese steel intensity peaked in 2009 and has been declining every year since — what we are witnessing in China and mining more broadly today is hardly a recent phenomenon.33 In any event, the implied consensus view on Chinese growth sees a metal growth multiplier tracking below one (see Exhibit 139).

We compare the IMF and consensus views on Chinese economic development and their implications for steel growth in Exhibit 140. The exhibit summarizes the sensitivity of the mining sector's value, which represents geared exposure to the iron ore price. It, in turn, represents geared exposure to Chinese steel demand, which then depends upon the overall growth of China's economy! Consequently, looking through these permutations and taking a view on how this will play out over the longer term is not an easy task, and we would interpret this as yet another reason for the miners to take a restrained approach to organic growth. The only ton that cannot be sold tomorrow is the one you sold today. Any doubt about the "natural" home for a ton suggests that it is better to leave it in the ground and wait till certainty is forthcoming. In 2010, Sam Walsh borrowed from Pliny the Elder and noted that "the only certainty is that nothing is certain."34 If 2010 was uncertain, then China is even more so today. We would therefore like to borrow a phrase from Pliny the Younger and note that one should "never do a thing concerning the rectitude of which you are in doubt." There is considerable doubt about the virtue of further iron ore growth.

33 And it is important to stress: peak steel intensity is only the year during which steel is embedded in the capital stock at the fastest rate. It does not mean that in absolute terms, the amount of steel embedded will decline. 34 MMC, December 2, 2010.

The Chinese GDP Growth Rate Consistent With Consensus (e) Steel Demand Is ~7%, or 1.3% Below the IMF and 0.5% Less Than the Official Government Target



AND LOVE THE ORE

Exhibit 136 Consensus Expectations for Chinese Steel Production Are Significantly Different from Those Implied by the IMF's GDP Deck

Source: CRU and Bernstein estimates and analysis. Exhibit 137 We Back Out the Scenario for China's GDP Evolution That Is Implied in the Steel

Growth Consensus; It Sees the Annual GDP Trend Growth of 7.1 vs. 8.5% Implied in the IMF Deck

Source: IMF, CRU and Bernstein estimates and analysis.

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

0

200

400

600

800

1000

1200


Cru

de

Ste

el P

rod

uct

ion

Gro

wth

Cru

de

Ste

el (M

t)Consensus Steel Production and Growth

2005A 2006A 2007A 2008A 2009A 2010A 2011A 2012A 2013E 2014E 2015E 2016E 2017E

IMF - April 2012 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.2% 8.2% 8.8% 8.7% 8.7% 8.6% 8.5%

IMF - April 2013 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 8.0% 8.2% 8.5% 8.5% 8.5%

Consensus 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 7.8% 7.6% 7.5% 7.2% 7.1%

6%

7%

8%

9%

10%

11%

12%

13%

14%

15%

Chinese GDP Growth — IMF vs. Implied Consensus



AND LOVE THE ORE 101

Exhibit 138 The Steel Intensity That Such a Trajectory Would Imply Is Markedly Lower Than the Scenario Consistent With the IMF GDP Deck

Source: WSA, IMF, CRU, Mitchell, Maddison and Bernstein estimates and analysis. Exhibit 139 Instead of a Recovery in the Metal to GDP Multiplier, We Would Continue to See

Declines

Source: WSA, IMF, CRU, Mitchell, Maddison and Bernstein estimates and analysis.

0

10

20

30

40

50

60

70

80

- 5 10 15 20 25 30 35 40 45 50 55

Ste

el I

nte

nsi

ty -

Kg

/'$00

0

Output - GDP per Capita '$000 (Real 2005 PPP)

Implied ConsensusSteel Intensity Forecast


-0.50

0.00

0.50

1.00

1.50

2.00

2.50

200

5A

200

6A

200

7A

200

8A

200

9A

201

0A

201

1A

201

2A

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

Chinese Steel Growth to GDP Growth Multiplier

IMF Consensus



AND LOVE THE ORE

Exhibit 140 Unsurprisingly, Given the Relative Immaturity of the Chinese Economy, Its Steel Growth Is Highly Leveraged (~4 to 1) to the Overall Economic Growth Rate

Source: WSA, IMF, CRU, Mitchell, Maddison and Bernstein estimates and analysis.

In addition to having visibility into the consensus view on steel production in China, we have some understanding of the Chinese cost structure (see Exhibit 141 and Exhibit 142). Moreover, we know the consensus price expectation (see Exhibit 143) — this price line shows the decline to ~US$80/t (real) that is currently discounted in equity valuations, albeit US$80/t appears to be factored into equity prices with no "fade" down. Rather, the market is indicating that any excess profits generated by the miners above this long-run equilibrium will simply be squandered on mistaken organic growth projects. Given this, we derive a consistent supply side response that is implicitly factored into the commodity price expectations (see Exhibit 144 through Exhibit 146). Over the next seven years, the market "expects" an increase in supply of ~550Mtpa delivered into the Chinese steel market, of which ~300Mtpa will be superfluous to the Chinese demand growth needs. The 300Mtpa of production will have to compete for space in the market through the displacement of high-cost incumbent tons. Consequently, we expect this to yield a decline in price along the trajectory assumed by consensus.

-6.0%

-4.0%

-2.0%

0.0%

2.0%

4.0%

6.0%

2012A 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

De

lta

Co

ns

ens

us

to I

MF

China GDP and Steel Growth IMF vs. Implied Consensus

GDP Steel

$80/t Iron Ore at ~7% Chinese GDP Is Only Maintained If ~550Mtpa of New Supply (330Mtpa Surplus) Also Come Onstream




Exhibit 141 Using Our Knowledge of the Structure of the Chinese Domestic Mining Industry, We Can Generate an Exemplar Cost Curve for the Industry as a Whole

Source: WSA, IMF, CRU, AME and Bernstein estimates and analysis. Exhibit 142 This Enables Us to Understand the Elasticity of Supply

Source: WSA, IMF, CRU, AME and Bernstein estimates and analysis.

0

20

40

60

80

100

120

140

160

180

200

0 200 400 600 800 1000 1200 1400 1600

Op

ex

(US

$/t)


Idealized Iron Ore Supply

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

Elas

tici

ty


Idealized Iron Ore Elasticity



AND LOVE THE ORE

Exhibit 143 Combining It With Consensus Steel Demand Expectations Yields Consensus Price Expectations…

Exhibit 144 …And Enables Us to Derive Consensus Supply Expectations

Source: Bloomberg L.P. Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.

Exhibit 145 Assuming the Consensus GDP Deck, Two-

Thirds of the Supply Expected to Come Onstream Is Not Needed…

Exhibit 146 …With Price Destructive Oversupply of ~300Mt by 2020

Source: WSA, IMF, CRU, Bloomberg L.P., Mitchell, Maddison and Bernstein estimates and analysis.


-

20

40

60

80

100

120

140

160

180

No

min

al (U

S$/

t)

Consensus Iron Ore Price Forecast

0

20

40

60

80

100

120

140

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

Su

pp

ly G

row

th o

f Iro

n O

re (M

t)

Implied Consensus Supply Expectations

Supply Required by Chinese Demand

Price Destructiv e Suppy in Excess of Demand

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

% of Excess Price Destructive Supply

Required Surplus

235

299

Total Iron Ore Supply by 2020 — Mt

Required Surplus




China Growth Scenarios

Using the model we built to back out Chinese GDP and supply side response that is consistent with consensus demand and price expectations, we can determine a consistent forecast for iron ore prices and steel demand growth across any GDP scenario for China. Our "bull" scenario takes the IMF's forecast from just over a year ago (April 2012), which sees the Chinese economy accelerate back up to a trend rate of 8.5%. This would absorb the ~300Mtpa of price destructive oversupply implied in the consensus scenario (mentioned earlier), resulting in a tight market in which the miners do not compete on volume and in which a trend price of US$140/t could eventuate out to 2020. A "middle" scenario takes the Chinese target of 7.5% GDP growth as a floor (despite the fact that Chinese GDP has a tendency to miss on the upside). This yields a price of US$120/t out to 2020. Our China "bear" scenario sees GDP decline to 6% and no supply coming out of the market (a true worst-case scenario). This results in an iron ore price testing the US$65/t mark over a multi-year period.35

The model we discussed in the previous chapter can be used to derive a self-consistent GDP, steel demand, supply reaction and iron ore price forecast for any particular GDP or price assumption. For example, using the consensus supply we derived in that, we assume that the market has a good grip on the supply side but sees demand (i.e., Chinese GDP) as uncertain, hence we can flex GDP growth assumptions and derive a new price forecast. Exhibit 147 summarizes forecast prices under the IMF GDP deck.36 The higher steel intensity of this GDP view against a fixed supply side response leads to higher prices. Consequently, under this scenario, prices are expected to average ~US$140/t out to 2020 — a far cry from the US$80/t that equities are currently discounting.

If it became clear that this price path were indeed to be realized, one could reasonably expect a supply side response, whereby more new growth projects would be approved. As we have stressed before, there is an inverse relationship between apparent and actual attractiveness of commodity markets. The more negative consensus expectations are, the less capital flows into the sector and thus the lower future volumes will be. Unsurprisingly, the lower future volumes in a commodity industry result in a higher future price. Conversely, the stronger consensus expectations are, the higher the capital inflows and the higher the volume growth, leading to a weaker counterfactual price environment. Once again, it is only by understanding the implications of counter-consensual and counter-cyclical behavior that sustainable value creation can be delivered in this industry. For the purposes of this analysis, we assume that the long lead time in new project development and approval will lead to the supply side changing more slowly than the demand side. Consequently, at least to a first order approximation, the resulting price line will be robust.

35 Our iron ore price forecast is discussed in greater detail in the "Our Iron Ore Price Forecast" chapter. 36 Ibid.

We Construct Bull, Middle, and Bear Scenarios for Chinese GDP and Supply Side Response to Bound the Possibilities for Iron Ore Prices Out to 2020 ($140/t Down to $65/t)

We Can Determine a Consistent Forecast for Iron Ore Prices and Steel Demand Growth Across Any GDP Scenario for China



AND LOVE THE ORE

Exhibit 147 We Can Use the Implied Consensus Supply Expectations to Derive Consensus Steel Demand Scenario and Iron Ore Price Line for Any Chinese GDP Deck Assumptions; Steel Intensity Is Geared to China's GDP and the Iron Ore Price Is Geared to Steel Intensity; Risk and Uncertainty Pervades This Sector


We construct three scenarios to stress test the iron ore price against different future development paths for China and against the supply view discussed earlier and in the "What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities?" chapter (see Exhibit 148 and Exhibit 149).

Scenario 1 — Back to the future: The IMF April 2012 view of GDP results in prices supported at an average ~US$140/t out to 2020. If the acceleration of the Chinese economy back up to a trend rate of 8.5% is to be believed, and this is the IMF base case, then that would require a corresponding acceleration in steel production. Given the relative immaturity of the Chinese economy, it is still at a very metal-intensive phase of its development. Consequently, accelerating growth requires acceleration in raw material consumption. The increasing productivity of the U.S. economy, which enabled it to grow output with no corresponding increase in metal intensity, occurred only post the conclusion of the capital stock build. China is still some way from this level (~4,500kg/capita versus 12,000kg/capita). For the Chinese economy to accelerate its growth without accelerating metal intensity, at this stage in its economic life, there would have to be an unprecedented increase in metal productivity. If we regard the implied supply expectations as being representative of what is most likely to eventuate in reality, such acceleration in steel intensity would mean that all new tons would be able to find a home. Instead of there being ~300Mtpa of price destructive oversupply, all of the ~550Mtpa of new supply would be able to satisfy true demand. So instead of having the emergence of the miners competing on volume, the market would stay tight and a trend price of US$140/t could eventuate out to 2020.

Scenario 2 — Ending with a whimper: Taking the Chinese base case of 7.5% GDP growth as a floor yields an average price of US$120/t out to 2020. There are, we believe, three factors that the Chinese government has at the heart of its industrial policy, all of which are focused on ensuring the maintenance of political stability. They are inflation (particularly for food), unemployment and corruption. The official GDP target for China is 7.5%. Historically, the official GDP target has been a floor rather than an average expectation, and is not subject to


Current SCB 33 42 46 76 81 147 168 128 131 138 142 145 139 135 123 113

Consensus 33 42 46 76 81 147 168 128 127 118 110 107 100 95 90 90

Consensus @ IMF GDP 33 42 46 76 81 147 168 128 138 142 152 160 157 150 129 114

0

20

40

60

80

100

120

140

160

180

Iro

n O

re —

FO

B A

us

Iron Ore Price Scenarios

Chinese GDP Growth Scenarios: 8.5%, 7.5% and 6% (and No Supply Reduction) Generate 2020 Iron Ore Prices of $140/t, $120/t and $65/t, Respectively




frequent revisions that might cause loss of face within the government (remember that during the period of an official 8% growth target, China averaged 10.6%). Moreover, this growth target is set to try and balance the competing objectives of avoiding inflation, while ensuring employment (clearly something easier in theory than in practice). The previous notwithstanding, this scenario shows the impact of a gradual fade to the official target levels of growth would actually be on China. The implication of this would be US$120/t iron ore out to 2020.

Scenario 3 — It's all over now baby blue: GDP revisions continue their descent down to 6%, and there is no curtailment of supply (a true bear case), resulting in prices testing US$65/t. This is for the true China bears. It shows the impact of continued downward revisions to Chinese growth and no curtailment of supply. It is worth noting that in a world of 6% growth, we believe that not only would many of the high-cost producers ($120-180/t) close, but a number of projects in the 550Mtpa of new supply would be canceled. However, a thorough bear scenario should leave all this supply intact and that is what we do. Under this world, the iron ore price could test the US$65/t level, just as it did in 2008; however, unlike 2008, we do not believe that the miners would be saved by the Chinese government stimulus. Consequently, a protracted period of low-price levels would persist. Troublingly, the iron ore majors would still be cash positive at these price levels, and so cutting existing tonnage as a form of self-help might still not occur. Again, the warning for us is clear...far better not to install capacity today than have to deal with trying to remove it if the world turns against you. (Not that it is a particularly good parallel, but Platinum used to be Anglo American's most profitable division; however, the good times have passed and trying to fix the industry through removing ounces and labor is proving virtually impossible.)

For each scenario, we provide the following data points: Overall Chinese GDP deck and how it differs from economic growth

expectations implied in consensus (see Exhibit 150, Exhibit 151, Exhibit 158, Exhibit 159, Exhibit 166 and Exhibit 167);

Steel intensity and production that would result in for a given GDP deck expectation and assuming that China will converge to U.S. productivity levels (see Exhibit 152, Exhibit 153, Exhibit 160, Exhibit 161, Exhibit 168 and Exhibit 169);

Steel growth as well as steel-growth-to-economic growth multiple (see Exhibit 154, Exhibit 155, Exhibit 162, Exhibit 163, Exhibit 170 and Exhibit 171);

Proportion of new supply that will be required to satisfy growing demand as well as the proportion that comes in excess of it, and would hence act to destroy the price (see Exhibit 156, Exhibit 164 and Exhibit 172); and

Resultant iron ore price forecast (see Exhibit 157, Exhibit 165 and Exhibit 173).

Please see "Appendix 2: Growth Scenarios" for more detail on the scenarios in this chapter and also the "What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities?" chapter.



AND LOVE THE ORE

Exhibit 148 In Order to "Stress Test" the Demand and Price Environment, We Construct Three Scenarios: (1) Slightly More Bullish Than the IMF (i.e., Back to the IMF April 2012 Figures); (2) One That Respects China's Official 7.5% GDP Growth Target; and (3) Significantly Below the Target and the Implied Consensus Growth Rates

Source: IMF and Bernstein estimates and analysis. Exhibit 149 The Resulting Price Paths Illustrate That the Risks for Long-Term Iron Ore Prices Are

to the Upside, Barring a Significant "Miss" on the Chinese Growth



IMF 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 8.0% 8.2% 8.5% 8.5% 8.5% 8.5% 8.1% 7.8%

Scenario 1: IMF 8.5% 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 8.2% 8.4% 8.7% 8.7% 8.7% 8.7% 8.3% 8.0%

Scenario 2: Official Target 7.5% 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 8.0% 8.0% 7.9% 7.7% 7.7% 7.7% 7.5% 7.5%

Scenario 3: China Bear 6% 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 7.6% 7.3% 7.0% 6.7% 6.3% 6.0% 6.0% 6.0%

5.0%

6.0%

7.0%

8.0%

9.0%

10.0%

11.0%

12.0%

13.0%

14.0%

15.0%

Chinese GDP Scenarios


Consensus 33 42 46 76 81 147 168 128 127 118 110 107 100 95 90 90

Scenario 1: IMF 8.5% 33 42 46 76 81 147 168 128 146 151 162 170 167 160 137 121

Scenario 2: Official Target 7.5% 33 42 46 76 81 147 168 128 137 134 127 125 122 117 106 104

Scenario 3: China Bear 6% 33 42 46 76 81 147 168 128 123 106 95 89 77 68 66 66

-

20

40

60

80

100

120

140

160

180

Iro

n O

re P

rice

—U

S$/

t

Iron Ore Price by GDP Scenario




Scenario 1 — "Back to the Future" A return to the world anticipated by the IMF in April 2012 would see a re-acceleration in the Chinese steel demand, with only 18% (rather than ~60%) of new supply being price destructive. Prices could stay at US$160/t for a sustained period of time.

Exhibit 150 Scenario 1 Uses a Slightly Stronger GDP

Deck Than the Current IMF Figures…

Exhibit 151 …And Nearly 2% Higher Than Consensus

Source: IMF and Bernstein estimates and analysis. Source: IMF and Bernstein estimates and analysis. Exhibit 152 To Support This Level of Growth, Steel

Intensity Would Need to Rise

Exhibit 153 It Would Enable an Accelerated Capital Stock Growth Rate in China



5.0%

6.0%

7.0%

8.0%

9.0%

10.0%

11.0%

12.0%

13.0%

14.0%

15.0%

Rea

l GD

P G

row

th

China GDP Deck

Scenario 1 Consensus

0.0%

0.2%

0.4%

0.6%

0.8%

1.0%

1.2%

1.4%

1.6%

1.8%

200

5A

200

6A

200

7A

200

8A

200

9A

201

0A

201

1A

201

2A

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

Scenario 1 GDP vs. Implied Consensus

40

45

50

55

60

65

70

2005

A

2006

A

2007

A

2008

A

2009

A

2010

A

2011

A

2012

A

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E

Kg

per

$'0

00

of G

DP

Chinese Steel Intensity


-

200

400

600

800

1,000

1,200

Cru

de

Ste

el (M

t)





AND LOVE THE ORE

Exhibit 154 This Would See Steel Growth Approach 10%...

Exhibit 155 ...And a Steel-to-GDP Ratio Well Above 1



Exhibit 156 Under This Scenario, Only 18% of New

Volume Is Price Destructive…

Exhibit 157 …Leading to a Continuation of High Prices Into the Medium Term



-5.0%

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

Chinese Crude Steel Production Growth


-0.5

0.0

0.5

1.0

1.5

2.0

2.5

Steel to GDP Growth


441

94

New Supply — Scenario 1 (Mtpa)

Required Additional Supply Price Destructiv e Supply

0

20

40

60

80

100

120

140

160

1802

005

A

200

6A

200

7A

200

8A

200

9A

201

0A

201

1A

201

2A

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

Iro

n O

re P

ric

e —

No

min

al —

US

$/t

FO

B A

us

Iron Ore Price





Scenario 2 — Ending With a Whimper If we respect a trend decline down to the government's target annual GDP growth of 7.5%, we would still see iron ore prices track some US$15/t above consensus.

Exhibit 158 Scenario 2 Assumes the Official Annual

GDP Growth Figure of 7.5%...

Exhibit 159 …Which Is Still Significantly Ahead of Consensus Expectations

Source: IMF and Bernstein estimates and analysis. Source: IMF and Bernstein estimates and analysis. Exhibit 160 Although It Can Be Achieved Without an

Upward Correction to Steel Intensity…

Exhibit 161 …It Still Anticipates an Above-Consensus Demand Growth…



5.0%

6.0%

7.0%

8.0%

9.0%

10.0%

11.0%

12.0%

13.0%

14.0%

15.0%

Rea

l GD

P G

row

th

China GDP Deck


0.0%

0.1%

0.2%

0.3%

0.4%

0.5%

0.6%

0.7%

2005

A

2006

A

2007

A

2008

A

2009

A

2010

A

2011

A

2012

A

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E


40

45

50

55

60

65

70

200

5A

200

6A

200

7A

200

8A

200

9A

201

0A

201

1A

201

2A

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

kg

pe

r $

'00

0 o

f G

DP



-

200

400

600

800

1,000

1,200

Cru

de

Ste

el (M

t)





AND LOVE THE ORE

Exhibit 162 …With Well-Supported Steel Growth... Exhibit 163 ...And a Declining Steel-to-GDP Ratio



Exhibit 164 Nevertheless, 200Mtpa of Supply Is Still

Unwarranted from a Value Perspective...

Exhibit 165 …Leading to Suppressed Yet Above-Consensus Iron Ore Prices



0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%



0.0

0.5

1.0

1.5

2.0

2.5

Steel to GDP Growth


334

200



0

20

40

60

80

100

120

140

160

180

2005

A

2006

A

2007

A

2008

A

2009

A

2010

A

2011

A

2012

A

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E

Iro

n O

re P

ric

e —

No

min

al

—U

S$

/t F

OB

Au

sIron Ore Price





Scenario 3 — "It's All Over Now Baby Blue" The third scenario sees all of the new supply growth as price destructive. It is consistent with China growing at 6% p.a. and would see iron ore prices test the US$65/t level.

Exhibit 166 Our Bear Case Assumes 6% GDP Growth

for China in the Long Term…

Exhibit 167 …Which Is ~1% Below the Current Implied Consensus Expectation

Source: IMF and Bernstein estimates and analysis. Source: IMF and Bernstein estimates and analysis. Exhibit 168 This Would Lead to a Significant Reduction

in Steel Intensity...

Exhibit 169 ...And Would See Stagnant Production Into the Medium Term



5.0%

6.0%

7.0%

8.0%

9.0%

10.0%

11.0%

12.0%

13.0%

14.0%

15.0%

Rea

l GD

P G

row

th

China GDP Deck


-1.2%

-1.0%

-0.8%

-0.6%

-0.4%

-0.2%

0.0%

2005

A

2006

A

2007

A

2008

A

2009

A

2010

A

2011

A

2012

A

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E


40

45

50

55

60

65

70

200

5A

200

6A

200

7A

200

8A

200

9A

201

0A

201

1A

201

2A

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

kg

pe

r $

'00

0 o

f G

DP



-

200

400

600

800

1,000

1,200

Cru

de

Ste

el (M

t)





AND LOVE THE ORE

Exhibit 170 Consequently, the Industry Would Effectively Go Ex Growth...

Exhibit 171 ...And Steel-to-GDP Ratio Would Fall Well Below 1



Exhibit 172 In This Scenario, Virtually All Volume

Growth Is Price and Value Destructive…

Exhibit 173 …Iron Ore Tests a US$65/t Floor



-5.0%

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%



-0.5

0.0

0.5

1.0

1.5

2.0

2.5

Steel to GDP Growth


6

528



0

20

40

60

80

100

120

140

160

18020

05A

2006

A

2007

A

2008

A

2009

A

2010

A

2011

A

2012

A

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E

Iro

n O

re P

rice

-N

om

inal

-U

S$/

t F

OB

Au

sIron Ore Price





Our Iron Ore Price Forecast

We believe that prices rise to the extent that marginal units of supply are called by demand and that prices fall to the extent that marginal units of supply are displaced. Because marginal supply determines price, the cost structure of an industry determines both the revenue and the margin generation of that industry. With this heuristic in place, we take inspiration from Fischer Black's "Business Cycles and Equilibrium," a book on business cycles and monetary policy,37 and construct a modified framework, replicating his approach to business cycles but extending the lead time between demand signals and the supply side response. One consequence of the long lead time is inherent cyclicality in the mining industry. Furthermore, this cyclicality ensures that the incentive price (the price at which a miner is incentivized to bring on new capacity) is seldom achieved, but is rather over- or under-shot. This means that the value generated by investment (either by the mining companies themselves or by investors into mining companies) is almost entirely a call on the ability to act counter-cyclically.

In previous chapters, we looked at scenario analysis conducted in mid-2Q:13. This chapter discusses our updated iron ore price forecast, distilled from the analytical frameworks of the previous chapters and their conclusions.

At the time of publication of this Blackbook, we predict $130/t (2013E) rising to a peak of $144/t (2016E). This is consistent with the IMF's current GDP forecast of 7.8% in 2013 rising to 8.5% in 2016 and average increase in supply of 88Mtpa 2013-16. By 2020, we expect iron ore prices of $115/t consistent with IMF forecasts for Chinese GDP growth 2017-20 and a further 28Mtpa of incremental supply in the market.

Our basic premise is that price signals refer to something in the real economy — namely, the marginal units of supply called to clear demand. Because marginal supply determines price, the cost structure of an industry determines both the revenue and the margin generation of the industry. We also note that the simple connection between a notional "oversupply" and falling prices or "undersupply" and rising prices embeds two implicit assumptions. The first is that price elasticity at the margin is high. The second is that cost escalation is low. In a world where these conditions do not hold, it is perfectly possible to have an oversupplied market with rising prices. We have to understand the structure of supply and demand at each point, and use that understanding to deduce the market clearing price. Prices rise to the extent that marginal units of supply are called by demand.

That call can originate from a number of sources, the most obvious of which is the increase in underlying economic growth. Other reasons include the depletion of existing non-marginal units of supply, as existing mines become exhausted.

Prices fall to the extent that marginal units of supply are displaced. The displacement can include both temporary falls in demand (e.g., through destocking and inventory management programs) and secular declines in demand (as a result of substitution or falling overall growth rates). The displacement can also occur as a consequence of the growth in non-marginal units of supply (which is typically described as an "oversupplied" industry).

While our interpretation of what supply and demand imply is perhaps a little different, the building blocks are traditional enough. The first element in our model is demand. We show how raw material consumption fits into the overall structure of economic output for a country. We do not believe that raw material consumption

37 Published in 1987.

We Apply a Fischer Black Framework Taken from Fischer Black's Work to Our Supply and Demand Analysis, Adapted to Long Lead Times in the Mining Sector; This Generates Our Counter-Consensus Positive View on Iron Ore Pricing

Our Basic Premise Is That Price Signals Refer to Something in the Real Economy



AND LOVE THE ORE

is an effect of economic growth — rather it is the cause of it. We understand the economy as primarily a mechanism for the conversion or processing of raw materials. Economic growth is then a measure of its relative intensity. This activity requires a scale of capital stock (machinery and infrastructure) capable of displacing human labor with the far more concentrated forms of energy contained in (primarily) fossil fuels. It is this displacement of labor by capital and high-intensity energy that generates high levels of productivity in the primary and secondary industry that we associate with developed economies. The displaced labor then moves into tertiary forms of value-add, leading to a society and an economy characterized by its service sector. However, the essential point is that the transition to a service sector cannot be occasioned without the development of a capital stock of sufficient scale.

With the previous discussion in mind, it is clear that China is currently at a point in its economic life where it is still installing this capital stock. Investment now accounts for 45% of Chinese GDP compared to 35% just 10 years ago (see Exhibit 174 and Exhibit 175). This, in turn, corresponds to the fact that over the last 10 years, secondary forms of economic activity in China have been the fastest-growing sector, outstripping growth in the tertiary forms of value-add by 1.4% per annum (see Exhibit 176). This implies a very significant capital stock formation in China. Moreover, it is the rapidity of this stock formation that stands behind the rapid acceleration in China's output over the last 10 years. We use this fundamental relationship between stock and output as the basis for our demand-side analysis (see Exhibit 177 through Exhibit 179). At this stage in China's development, the economy remains very metal intensive. Consequently, any re-acceleration in GDP back to the trend levels forecast by the IMF (8.5%) or OECD (8.4%) must translate into a corresponding acceleration in metal demand. In developed economies, accelerating GDP means accelerating consumption. However, the composition of the Chinese economy is such that consumption alone cannot account for economic growth. Rather, it must be supplemented with investment, which, in turn, means metal.

Exhibit 174 There Has Been a Significant Change in the

Structure of the Chinese Economy Over the Last 10 Years...

Exhibit 175 ...With Consumption Losing Out to Investment; Now Investment Accounts for Around Half of All Activity

Source: Bloomberg L.P. and Bernstein estimates and analysis. Source: Bloomberg L.P. and Bernstein estimates and analysis.

45%

16%

35%

2%2%

Composition of Chinese GDP (2002)

Household Consumption Govt Consumption

Fixed Capital Formation Inventory Formation

Net Exports

35%

14%

46%

2% 3%

Composition of Chinese GDP (2012)

Household Consumption Govt Consumption

Fixed Capital Formation Inventory Formation

Net Exports

At This Stage in China's Development, Its Economy Remains Metal Intensive; Any Re-Acceleration in China's GDP Back to Trend Levels Forecast by the IMF (8.5%) or OECD (8.4%) Must Translate Into a Corresponding Acceleration in Metal Demand




Exhibit 176 Secondary Forms of Economic Activity Have Grown on Average 1.4% Faster Than the Tertiary Forms Over the Course of China's Industrialization (10.7% vs. 9.3%)

Source: Bloomberg L.P. and Bernstein estimates and analysis. Exhibit 177 However, the Net Result of All of This Activity Is That China Is on Trend to Catch Up

With the U.S. in Terms of the Overall Steel Productivity

Source: WSA, Mitchell, Maddison, Bloomberg L.P., IMF and Bernstein estimates and analysis.

0

2

4

6

8

10

12

14

16

18Ju

n-99

Nov

-99

Apr

-00

Sep

-00

Feb-

01

Jul-0

1

Dec

-01

May

-02

Oct

-02

Mar

-03

Aug

-03

Jan-

04

Jun-

04

Nov

-04

Apr

-05

Sep

-05

Feb-

06

Jul-0

6

Dec

-06

May

-07

Oct

-07

Mar

-08

Aug

-08

Jan-

09

Jun-

09

Nov

-09

Apr

-10

Sep

-10

Feb-

11

Jul-1

1

Dec

-11

May

-12

Oct

-12

Mar

-13

Yo

Y G

DP

Ch

ang

e (%

)Chinese GDP Growth Composition

Primary Secondary Tertiary

-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

- 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0

Kg

of S

teel

per

Cap

ita


Capital Stock of Steel — China Forecast vs. USA




AND LOVE THE ORE

Exhibit 178 China Currently Has ~35% of the Steel Stock of the U.S.; We Expect It to Rise to ~69% by 2020

Source: WSA, Mitchell, Maddison, Bloomberg L.P., IMF and Bernstein estimates and analysis. Exhibit 179 We Use This Relationship Between Steel Stock and Output to Link the Steel Intensity

(i.e., Rate of Steel Consumption) Between Different Periods in a Country's Economic Development

Source: WSA, Mitchell, Maddison, Bloomberg L.P., IMF and Bernstein estimates and analysis.

However, there is more than just the steel demand figure that needs to be considered. Specifically, it has to be put in the context of an overarching view on the development of the Chinese economy. Would the figure that we generate for steel demand imply, for example, an increase or a decrease in the fixed-asset investment (FAI) ratio for China? Exhibit 180 shows that the increase in steel demand that we forecast is commensurate with a decline in the FAI ratio in China back down to ~38% of GDP. As such, it corresponds to the sought-after "rebalancing" of the Chinese economy. In order to explicate the commensurability of increasing steel use with falling FAI, it is worth bearing in mind that the high

0

100

200

300

400

500

600

700

800

900

1000

0%

10%

20%

30%

40%

50%

60%

70%

80%

2005

A

2006

A

2007

A

2008

A

2009

A

2010

A

2011

A

2012

A

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E

Fin

ish

ed

Ste

el D

ema

nd

(M

t)

Ch

ines

e C

ap

ital

Sto

ck

as

% o

f US

AChinese Finished Steel Demand and Capital Stock

Finished Steel Demand Capital Stock as % of USA

0

10

20

30

40

50

60

70

80

- 5 10 15 20 25 30 35 40 45 50 55

Ste

el I

nte

nsi

ty —

Kg

/$'0

00



China Actual China Forecast (IMF GDP) SCB Trend




FAI ratio in China has very little to do with raw material consumption per se. An analysis of the most important raw materials in China establishes that only about 12% of the FAI is tied to raw material consumption (see Exhibit 181), even given the most aggressive assumption that 100% of metal demand in China occurs as investment rather than consumption. Our forecast for steel use sees this 12% number declining over time (see Exhibit 182). Furthermore, it is clear that only three metals really matter to the Chinese economy considered as a whole: steel, copper and aluminum. Between them, they account for ~80% of the FAI attributable to metal consumption (see Exhibit 183 and Exhibit 184).

Exhibit 180 Such a Trajectory of Steel Use Implies FAI Falling from 48% to 38% of GDP, in Effect

Achieving the Policy Aim of Rebalancing the Chinese Economy

Source: WSA, Mitchell, Maddison, Bloomberg L.P., IMF and Bernstein estimates and analysis. Exhibit 181 However, the Vast Majority of the Investment in China Relates Not to Raw Material

Consumption Per Se, But Rather the Capitalization of Labor Associated With That Investment

Source: Bloomberg L.P., AME, CRU, Wood Mackenzie and Bernstein estimates and analysis.

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

0

5,000

10,000

15,000

20,000

25,000

30,000

2005

A

2006

A

2007

A

2008

A

2009

A

2010

A

2011

A

2012

A

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E

Composition of Chinese GDP FAI and Consumption

FAI in GDP Non FAI in GDP FAI as Proportion of Total

12%

88%

Composition of FAI in 2012

Raw Materials in FAI Non-Raw Materials in FAI



AND LOVE THE ORE

Exhibit 182 Under Our Scenario for Raw Material Consumption (With Constant Prices), the Raw Material Component of FAI Falls from 12.4% (Historical Average) to 8.9% Over the Forecast Period

Source: Bloomberg L.P., AME, CRU, Wood Mackenzie and Bernstein estimates and analysis. Exhibit 183 Steel Is by Far the Most Important Raw Material Component of FAI; Copper and

Aluminum Are Also Significant


4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

16.0%20

05A

2006

A

2007

A

2008

A

2009

A

2010

A

2011

A

2012

A

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E

Raw Material Consumption Contribution to FAI

61%

6%

10%

2%2%1%

18%

Raw Material Breakdown in Chinese FAI

% Steel of FAI % Aluminum of FAI % Copper of FAI % Zinc of FAI % Nickel of FAI % Lead of FAI % Cement of FAI




Exhibit 184 The Proportion of Raw Materials in FAI Has Stayed Very Stable Over Time


The real issue with the FAI ratio is not raw material consumption but the capitalization of labor. Exhibit 185 shows how the labor pool in China has moved over time as well as our forecast evolution of it going forward. Exhibit 186 shows a continuation of the trend rates of loss of labor from agriculture, as the de-agrarianization of China continues. We also extend the trend rate adding labor to the service sector. This leaves the secondary sector employment as the balancing item in the labor pool in the context of an overall contraction in the workforce in China from 2014. Consequently, this sees the secondary sector experiencing significant labor losses by ~2016 and is again in line with economic "rebalancing."

Exhibit 187 shows, for interest, the process of capital stock formation and de-agrarianization in China. The capital stock in China is on track to be some 15% larger than that in the U.S. before the transition to a U.S.-style service sector economy is complete. The main takeaway from this analysis, however, is shown in Exhibit 188. It illustrates how rising steel demand with falling FAI in China can be squared through increasing labor productivity in the secondary sector. The increases in steel demand are more than offset by the declines in the labor employed in the secondary sector. It is this productivity increase that also corresponds to the increase in the overall monetary value of the Chinese economy, with more goods circulating ever more efficiently through society.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2005

A

2006

A

2007

A

2008

A

2009

A

2010

A

2011

A

2012

A

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E

Composition of Raw Materials in FAI Over Time

% Steel of FAI % Aluminum of FAI % Copper of FAI % Zinc of FAI % Nickel of FAI % Lead of FAI % Cement of FAI



AND LOVE THE ORE

Exhibit 185 Against This Picture of Raw Material Use, the Increase in the Importance of Investment in China Is Due to the Movement of Labor Into Tertiary Forms of Activity

Source: Bloomberg L.P., World Bank, NBS and Bernstein estimates and analysis.

Exhibit 186 Against a Backdrop of Continued De-Agrarianization, a Declining Labor Pool and a Continued Move Into Tertiary Forms of Employment, the Component of the Labor Force in Secondary Forms Must Decline Over Time

Source: Bloomberg L.P., World Bank, NBS and Bernstein estimates and analysis.

0

100

200

300

400

500

600

700

800

900

2005

A

2006

A

2007

A

2008

A

2009

A

2010

A

2011

A

2012

A

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E

Eco

no

mic

ally

Act

ive

Pe

ople

(mn)

Primary Labor Secondary Labor Tertiary Labor

-8.0%

-6.0%

-4.0%

-2.0%

0.0%

2.0%

4.0%

6.0%

8.0%

20

06A

20

07A

20

08A

20

09A

20

10A

20

11A

20

12A

20

13E

20

14E

20

15E

20

16E

20

17E

20

18E

20

19E

20

20E

Change in Chinese Labor Force Structure

Change in Primary Labor Change in Secondary Labor Change inTertiary Labor




Exhibit 187 The Trajectory of De-Agrarianization Has China Requiring 15% More Steel Capital Stock Than the U.S. to Effect the Transition to an Economy Dominated by the Service Sector

Source: Bloomberg L.P., World Bank, NBS and Bernstein estimates and analysis. Exhibit 188 This Deceleration in Steel Demand and Change in Composition of the Workforce

Imply a Significant Increase in Steel Use Productivity; If Steel Productivity Were Held Flat, There Would Be Significant Upside to Steel Demand

Source: Bloomberg L.P., AME, CRU, Wood Mackenzie, IMF, World Bank, NBS and Bernstein estimates and analysis.

R² = 0.9987

0

5

10

15

20

25

30

35

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 110%

Imp

ort

anc

e o

f Ag

ric

ult

ure

in C

hin

a R

ela

tive

to U

SA

Capital Stock of China Relative to USA

De-Agrarianization of China Implied in Capital Stock Accumulation

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

0

50

100

150

200

250

300

200

5A

200

6A

200

7A

200

8A

200

9A

201

0A

201

1A

201

2A

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E P

rod

uct

ivit

y o

f Sec

on

dar

y L

ab

or

—kg

Ste

el/C

apit

a

Lab

or

in t

he

Sec

on

da

ry S

ect

or

Steel Use Productivity

Secondary Labor Steel Productiv ity



AND LOVE THE ORE

On the supply side, we use our analysis of the main mining companies (Vale, Rio Tinto, BHP Billiton and Anglo) as well as third-party data for Fortescue Metals Group (ticker FMG; not covered) to derive the supply side response from the incumbent producers. We supplement this with an estimate of iron ore supply from new entrants and other non-major sources of supply (see Exhibit 190 through Exhibit 192). Then, we estimate what we believe is a representative proxy for the cost structure of the iron ore industry landed China. Clearly, this involves the conversion of a discrete set of individual mines into a continuous output function for the iron ore industry as a whole. However, given the inherent uncertainty in any price forecast, we believe that this step is warranted. We then estimate how the cost structure would evolve if price equilibrium were to be maintained, i.e., if supply changes were to exactly equal demand changes. We supplement this with an estimate of how the structure would change as a consequence of the cost escalation that we continue to see in the Chinese domestic industry. In this regard, it is useful to quickly summarize the two types of movements within the cost structure of a commodity market. Rotations of the cost curve — change margin but may or may not change price.

This change implies differential in cost escalation within the industry and hence a change in competitive position, privileging some players and disadvantaging others. This creation of relative advantage generates margin.

Translations — change price but leave margin unchanged. Translations of the cost curve imply that all players have been impacted equally by any new industry development. Consequently, relative competitive position is unchanged and the change has no impact on the margin.

Margin in mining is created by geological and mining method differentiation. This differentiation then translates into cost-curve rotations, which lead to margin increases and (assuming that the change has not taken place through an unwarranted expenditure of capital) value. For example, in iron ore this has arisen as a consequence of the difference between truck and shovel operations in Australia (mining high-grade hematite) versus wheelbarrow and muscle operations in China (mining low-grade magnetite). We show the impact of the cost-curve changes in Exhibit 189.

Exhibit 189 Against This Demand Environment, We Look at Cost Curves of Iron Ore Landed China; the New Low-Cost Supply Additions Rotate the Cost Curve Downwards; Cost Inflation Translates the Cost Curve Upwards; It Is the Interplay Between Cost Curve Translation and Rotation That Generates Margin and Ultimately Price

Note: 2016 and 2020 numbers are Bernstein estimates.

Source: CRU, AME and Bernstein estimates and analysis.

20

40

60

80

100

120

140

160

180

0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000 1050 1100 1150

US

$/t —

FO

B A

us

—N

om

inal

Cumulative Volume (Mt)

Smoothed Chinese Cost Curves Evolution

2012 2016 2020

We Look at the Supply Response from the "Big 4+1" to Understand How the Cost Curve (Landed China) Will Change Over Time; This Cost Curve Dynamic Drives the Expected Price Performance




Exhibit 190 Our Analysis of the Major Miners Informs Our Supply Side Projections

Source: CRU, corporate reports and Bernstein estimates and analysis. Exhibit 191 We Allow for Significant Growth from the

Juniors

Exhibit 192 FMG Appears to Have the Most Aggressive Growth Plans

Source: CRU, corporate reports and Bernstein estimates and analysis. Source: CRU, corporate reports and Bernstein estimates and analysis.

-

200

400

600

800

1,000

1,200

1,400

1,600

2005

A

2006

A

2007

A

2008

A

2009

A

2010

A

2011

A

2012

A

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E

Mt I

ron

Ore

"Big Five" Iron Ore Supply

Rio BHP FMG Vale Anglo

- 50 100 150 200 250

Other

BHP

Vale

Rio

FMG

Anglo

Iron Ore (Mt)

Nw Iron Ore Supply to 2020E

0.0% 2.0% 4.0% 6.0% 8.0% 10.0%

FMG

BHP

Rio

Anglo

Vale

Volume Growth CAGR to 2020E



AND LOVE THE ORE

Having calculated a supply and a demand response, we look at how the balance in the iron ore market will change over time. It should be stressed that this balance is purely notional. Iron ore is not copper — we do not expect to see 200mt (say) of "spare" iron ore on board a queue of capesize ships outside Qingdao. Rather, there will be changes to system-wide utilization rates, and these changes will be reflected in the corresponding cost structures of the various steps along the steel value chain. These balances are shown in Exhibit 193 for us and Exhibit 194 for consensus. Exhibit 195 then compares our estimates to consensus.

We are more bullish than consensus on demand — a positive for price. Consequently, either consensus has not understood what it will take to re-accelerate the Chinese economy correctly, or the IMF and OECD have underestimated the Chinese dislike of paying for raw materials. We also remain ahead of consensus on the supply side — a negative for price. At the time of our launch in September 2012, we expected a low-cost supply response of over one billion tons by 2020 and a "catastrophic" real price collapse down to ~US$75/t. We are now more inclined to believe that a degree of capital discipline is entering the industry, and that radical overbuild of mining capacity will be avoided.

However, supply and demand are only one component of any price analysis — the other is cost. If the cost structure of an industry is changing faster than the rate at which marginal units of supply are either called or displaced, and if the elasticity of supply at the margin is low, then oversupply can be accompanied by rising prices and vice versa. Despite the simple continuous output function that we have used in this analysis, we believe that there is a good reason to believe that the elasticity of supply in China (once one is at a high-enough cost level) is actually relatively flat. Once you are prepared to mine below, say, 15% Fe ROM, there is an abundance of such material that can be exploited even in China. And if the method of exploitation is similar, then the cost structure of the industry at the margin ought to be similar as well, leading to a low elasticity of supply. However, this would represent an upside to our price forecast, as it is already ahead of consensus. Consequently, we thought it best not to model this effect in our current version of the analysis. However, the change in elasticity at the margin may be structured — cost escalation in China is real enough (see Exhibit 196).

Putting all of these elements together generates our counter-consensus iron ore price forecast (see Exhibit 197 and Exhibit 198). Moreover, we are able to disaggregate the difference between our view and the view implicit in the consensus price forecast, as shown in Exhibit 199. Approximately half of the variance to consensus is attributable to our understanding of cash cost escalation in the Chinese mines, and the other half is due to a difference in supply and demand balance. Exhibit 200 through Exhibit 203 provide the full details of this analysis.

Three Variables Need to Be Considered in Any Price Forecast — Supply, Demand and Cost; However, We Believe That Cost Is Commonly Omitted Through a Desire Not to Confuse Nominal and Real Drivers of Price Appreciation; Yet Differential Real Cost Escalation Is a Key Source of Value Creation in Mining




Exhibit 193 We Then Look at Changes to the Supply and Demand Situation to Imply a Change in the Overall "Balance" of the Iron Ore Industry...

Source: CRU, corporate reports and Bernstein estimates and analysis. Exhibit 194 ...Which Is, Naturally Enough, Different from Consensus

Source: CRU, corporate reports and Bernstein estimates and analysis.

(150)

(100)

(50)

-

50

100

150

2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

Iro

n O

re (M

t)SCB S-D Iron Ore Balance

SCB Change in Demand SCB Change in Supply SCB Change in Balance

(150)

(100)

(50)

-

50

100

150

2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

Iro

n O

re (M

t)

Consensus S-D Iron Ore Balance

Consensus Change in Demand Consensus Change in Supply Consensus Change in Balance



AND LOVE THE ORE

Exhibit 195 We Expect a Stronger Demand Environment and a Stronger Supply Side Response Than Consensus

Source: CRU, corporate reports and Bernstein estimates and analysis. Exhibit 196 A Critical Difference Is in the Impact of Expected Continued Cost Escalation on a

Non-Static Chinese Iron Ore Industry

Source: NBS and Bernstein estimates and analysis.

(150)

(100)

(50)

-

50

100

150

2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

Iro

n O

re (M

t)Consensus vs. SCB Comparison S-D Iron Ore Balance

Comparison Change in Demand Comparison Change in Supply Comparison Change in Balance

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

Chinese Iron Ore Cost Escalator




Exhibit 197 It Is a Combination of the Differences in Supply and Demand Forecasts as Well as Expected Real Cost Escalation in China That Gives Rise to Our Above-Consensus Price Forecast

Source: NBS, Bloomberg L.P., CRU and Bernstein estimates and analysis. Exhibit 198 Our Iron Ore Price Forecast Is Consistent With the IMF's Current GDP Forecast of

7.8% in 2013 Rising to 8.5% in 2016 and Average Increase in Supply of 88Mtpa 2013-16E


-

20

40

60

80

100

120

140

160

2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

Iro

n O

re (

US

$/t

)

Resulting Price Path and Comparison to Consensus

Price Dif f erence SCB Price Forecast Consensus Price Forecast


Iron Ore — US$/t 109 129 130 137 141 144



*as of June 30, 2013



AND LOVE THE ORE

Exhibit 199 Roughly 50% of the Upside of Our Forecast vs. Consensus Is Attributable to Differences in Our Cost Escalation Assumptions and 50% Due to S-D Differences

Source: NBS, Bloomberg L.P., CRU and Bernstein estimates and analysis.

(5)

-

5

10

15

20

25

30

2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

Iro

n O

re (U

S$/

t)

Explanation of Price Differential vs. Consensus

Cost Supply & Demand


E

UR

OP

EA

N M

ET

AL

S &

MIN

ING

: A S

TR

AN

GE

LO

VE

— H

OW

I LE

AR

NE

D T

O S

TO

P W

OR

RY

ING

AN

D L

OV

E T

HE

OR

E 131

Exhibit 200 The First Element of an Iron Ore Price Forecast, in Our View, Is the Calibration of a Consistent View on Steel Intensity and Steel Productivity


Chinese Steel Intensity Analysis 2005A 2006A 2007A 2008A 2009A 2010A 2011A 2012A 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

China Population m 1,298 1,304 1,311 1,317 1,324 1,329 1,335 1,340 1,346 1,352 1,357 1,361 1,364 1,368 1,372 1,375

China GDP Growth % 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 7.8% 8.0% 8.2% 8.2% 8.2% 8.2% 8.1% 7.8%

China GDP (Real 2005 PPP$) $bn 5,364 6,044 6,900 7,565 8,262 9,125 9,973 10,751 11,585 12,511 13,537 14,647 15,848 17,148 18,537 19,984

China GDP/Capita (Real 2005 PPP$) $/Capita 4,133 4,634 5,265 5,744 6,242 6,866 7,472 8,021 8,607 9,257 9,974 10,763 11,615 12,535 13,515 14,532

Chinese Steel Intensity kg/'000$ 65.5 63.1 61.8 57.8 66.7 63.3 63.2 61.5 59.4 60.5 60.6 59.2 57.0 54.5 51.2 46.9

China Finished Steel Demand Mt 351 381 426 437 551 578 630 662 688 756 820 866 903 935 949 938

China Finished Steel Demand Growth % 25.2% 8.5% 11.8% 2.5% 26.0% 4.9% 9.0% 5.0% 4.0% 10.0% 8.4% 5.6% 4.3% 3.5% 1.6% -1.2%

Chinese Self-Sufficiency in Steel % 98.9% 107.6% 110.7% 108.8% 99.8% 103.9% 103.9% 103.9% 104.2% 104.6% 104.9% 105.2% 105.5% 105.7% 106.0% 106.4%

Chinese Finished Steel Production Mt 348 410 472 475 550 600 655 687 717 792 861 912 953 988 1,007 997

Finished Steel to Crude Steel Ratio % 101.6% 102.2% 103.7% 105.2% 104.3% 104.4% 104.4% 104.4% 104.8% 105.1% 105.4% 105.7% 105.9% 106.2% 106.4% 106.6%

China Crude Steel Production Mt 353 419 489 500 574 627 683 717 751 832 907 964 1,010 1,049 1,071 1,063

China Crude Steel Production Growth % 24.9% 18.7% 16.7% 2.3% 14.6% 9.3% 9.0% 5.0% 4.3% 10.4% 8.7% 6.0% 4.5% 3.7% 1.9% -0.9%

Scrap Consumed in EAF Mt 43 46 46 48 58 64 70 74 79 86 90 95 99 102 105 107

Scrap Consumed in BOF Mt 18 22 36 53 44 59 74 86 100 117 129 143 156 168 178 187

Iron Ore Demand Mt 538 648 751 739 870 930 999 1,034 1,061 1,168 1,277 1,347 1,402 1,448 1,465 1,433

Chinese Steel Capital Stock Mt 2,625 2,954 3,321 3,692 4,169 4,663 5,200 5,757 6,330 6,960 7,641 8,355 9,091 9,844 10,596 11,322

Chinese Steel Capital Stock/Capita kg/Capita 2,023 2,265 2,534 2,803 3,150 3,508 3,896 4,295 4,703 5,149 5,630 6,139 6,663 7,196 7,725 8,233

Steel Growth to GDP Growth Ratio - 2.2 0.7 0.8 0.3 2.8 0.5 1.0 0.6 0.5 1.2 1.0 0.7 0.5 0.4 0.2 -0.2

For the exclusive use of JA

SO

N LA

PO

RT

E at P

ER

RY

CA

PIT

AL on 16-Jul-2013

132 E

UR

OP

EA

N M

ET

AL

S &

MIN

ING

: A S

TR

AN

GE

LO

VE

— H

OW

I LE

AR

NE

D T

O S

TO

P W

OR

RY

ING

AN

D L

OV

E T

HE

OR

E

Exhibit 201 We Calibrate This Against the Overall Composition of Economic Activity and the Labor Force in China


Chinese Macro & Raw Material Summary 2005A 2006A 2007A 2008A 2009A 2010A 2011A 2012A 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E


China GDP Growth (Real) % 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 7.8% 8.0% 8.2% 8.2% 8.2% 8.2% 8.1% 7.8%



FAI as % of GDP % 42.1% 43.0% 41.7% 44.0% 48.2% 48.2% 47.6% 47.0% 47.1% 46.8% 46.0% 44.9% 43.3% 41.6% 39.7% 37.8%

Inflation Index % 100.0% 103.2% 106.2% 108.6% 109.5% 111.0% 113.4% 115.4% 117.6% 120.2% 123.0% 125.9% 129.0% 132.0% 135.2% 138.4%

China GDP (Nominal PPP$) $bn 5,364 6,240 7,330 8,214 9,049 10,128 11,306 12,406 13,623 15,039 16,647 18,443 20,441 22,641 25,055 27,651

Nominal GDP Growth % 16.3% 17.5% 12.1% 10.2% 11.9% 11.6% 9.7% 9.8% 10.4% 10.7% 10.8% 10.8% 10.8% 10.7% 10.4%

FAI Contribuition to GDP $bn 2,258 2,681 3,059 3,618 4,366 4,884 5,377 5,830 6,418 7,036 7,664 8,272 8,856 9,415 9,952 10,456

Steel Consumption Mt 351 381 426 437 551 578 630 662 688 756 820 866 903 935 949 938

Steel per $'000 FAI kg/'$000 156 142 139 121 126 118 117 113 107 108 107 105 102 99 95 90

Steel Price — RMB CNY/t 4,234 3,935 4,286 5,014 3,728 4,256 4,673 3,999 4,309 4,309 4,309 4,309 4,309 4,309 4,309 4,309

RMB Exchange Rate CNY:USD 8.19 7.97 7.61 6.95 6.83 6.77 6.46 6.31 6.40 6.45 6.51 6.57 6.64 6.69 6.75 6.81

Steel Price — USD US$/t 517 494 563 721 546 629 723 634 673 668 662 656 649 644 638 633

Steel Cost of FAI $bn 182 188 240 315 301 363 456 419 463 505 543 568 587 602 606 593

Total Raw Material Consumption in FAI US$bn 267 316 408 496 480 602 745 683 739 795 844 881 908 930 942 936

Non-Raw Material FAI Cost (Labor) US$bn 1,991 2,365 2,652 3,122 3,886 4,282 4,632 5,147 5,678 6,241 6,820 7,392 7,948 8,485 9,010 9,521

Raw Material as % of FAI % 12% 12% 13% 14% 11% 12% 14% 12% 12% 11% 11% 11% 10% 10% 9% 9%

Chinese Economically Active Population m 779 763 765 770 775 784 785 788 790 791 790 789 786 782 778 775

Chinese Employed Population m 758 764 753 756 758 761 764 767 769 770 769 768 765 761 758 754

Unit Value of Labor in FAI US$/person 11,010 12,302 13,136 15,191 18,434 19,605 20,546 22,221 24,048 26,141 28,483 31,057 33,875 36,929 40,227 43,719

Growth in Labor Cost in FAI % 11.7% 6.8% 15.6% 21.3% 6.4% 4.8% 8.1% 8.2% 8.7% 9.0% 9.0% 9.1% 9.0% 8.9% 8.7%

Steel Use Productivity kg/capita 1,943 1,983 2,111 2,126 2,613 2,645 2,795 2,856 2,913 3,168 3,426 3,640 3,850 4,068 4,238 4,305


SO

N LA

PO

RT

E at P

ER

RY

CA

PIT

AL on 16-Jul-2013

E

UR

OP

EA

N M

ET

AL

S &

MIN

ING

: A S

TR

AN

GE

LO

VE

— H

OW

I LE

AR

NE

D T

O S

TO

P W

OR

RY

ING

AN

D L

OV

E T

HE

OR

E 133

Exhibit 202 We Look at How the Cost Curve in China Would Change If There Were to Be No Impact on Price; We Then Calculate the Impact of a Notional "Over Supply" or "Under Supply" on That Cost Curve, Particularly Relative to Consensus Expectations


Iron Ore Cost Curve Analysis 2005A 2006A 2007A 2008A 2009A 2010A 2011A 2012A 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

Consensus Iron Ore Forecast US$/t 33 42 46 76 81 147 168 128 127 118 110 107 100 95 90 90

Consensus Crude Steel Production Mt 353 419 489 500 574 627 683 717 751 794 828 858 885 914 943 973

Assumed Consensus Scrap Consumed in EAF Mt 43 46 46 48 58 64 70 74 79 86 90 95 99 102 105 107

Assumed Consensus Scrap Consumed in BOF Mt 18 22 36 53 44 59 74 86 100 117 129 143 156 168 178 187

Implied Consensus Iron Ore Demand Mt 538 648 751 739 870 930 999 1,034 1,061 1,098 1,133 1,154 1,175 1,200 1,232 1,269

a (C(Q) = a + b*Q^n) - 40 40 40 40 40 40 40 40 40

b (C(Q) = a + b*Q^n) - 2.85E-06 2.67E-06 2.45E-06 2.27E-06 2.16E-06 2.07E-06 1.96E-06 1.84E-06 1.71E-06

n (C(Q) = a + b*Q^n) - 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5

Qbase Mt Iron Ore 27 37 34 22 20 26 31 38

Cumulative Qexcess Mt Iron Ore 49 98 143 166 209 248 291 299

Qexcess Mt Iron Ore 49 49 44 23 43 38 43 9

Consensus to Calculated Crude Steel Mt 0.4 37.8 78.8 105.6 124.2 135.4 127.6 89.5

Chinese Iron Ore Cost Escalation % 9.6% 12.9% 6.4% 2.4% 3.5% 4.9% 4.4% 4.1%

a' (C'(Q) = a' + b'*Q^n') - 40 44 49 53 54 56 59 61 64

b' C'(Q) = a' + b'*Q^n') - 2.85E-06 2.67E-06 2.45E-06 2.27E-06 2.16E-06 2.07E-06 1.96E-06 1.84E-06 1.71E-06

n' C'(Q) = a' + b'*Q^n') - 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5

Iron Ore Production Available for Export 2005A 2006A 2007A 2008A 2009A 2010A 2011A 2012A 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

Rio Mt 159 168 179 193 218 239 245 253 267 316 364 378 392 377 377 377

BHP Mt 114 115 120 136 137 149 174 187 209 230 238 239 249 292 324 324

FMG Mt 15 33 42 48 64 101 110 112 113 116 125 130 130

Vale Mt 234 264 303 302 238 308 323 320 308 343 375 384 397 412 441 452

Anglo Mt 31 31 32 37 45 47 46 49 49 48 59 73 73 73 73 73

Big Five Mt 538 579 634 684 670 785 835 873 935 1,048 1,147 1,186 1,227 1,278 1,344 1,355

Change from Big Five Mt 62 113 100 39 41 51 67 11

Change from Other Sources Mt 16 (7) 6 22 45 36 34 52

SCB Total Change in Potential Supply Mt 78 106 106 61 86 86 100 63

Consensus Supply Growth Mt 77 86 78 45 64 64 74 46

SCB vs. Consensus Supply Mt 1 19 27 16 22 22 26 16


SO

N LA

PO

RT

E at P

ER

RY

CA

PIT

AL on 16-Jul-2013

134 E

UR

OP

EA

N M

ET

AL

S &

MIN

ING

: A S

TR

AN

GE

LO

VE

— H

OW

I LE

AR

NE

D T

O S

TO

P W

OR

RY

ING

AN

D L

OV

E T

HE

OR

E

Exhibit 203 This Then Drives Our Price Line


S/D Balance & Price 2005A 2006A 2007A 2008A 2009A 2010A 2011A 2012A 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

SCB Chinese Demand Mt 538 648 751 739 870 930 999 1,034 1,061 1,168 1,277 1,347 1,402 1,448 1,465 1,433

SCB Change in Chinese Demand Mt 27 107 109 71 54 46 17 (32)

SCB Change in Supply Mt 78 106 106 61 86 86 100 63

SCB Balance Mt (51) 1 3 10 (32) (40) (83) (95)

Consensus Chinese Demand Mt 538 648 751 739 870 930 999 1,034 1,061 1,098 1,133 1,154 1,175 1,200 1,232 1,269

Consensus Change in Chinese Demand Mt 27 37 34 22 20 26 31 38

Consensus Change in Supply Mt 77 86 78 45 64 64 74 46

Consensus Balance Mt (49) (49) (44) (23) (43) (38) (43) (9)

Volume Required for Price Equilibrium ex. Cost Inflation Mt 27 37 34 22 20 26 31 38

SCB Relative Supply Position vs. 2012 Equilibrium Condition Mt 1,010 1,049 1,086 1,118 1,106 1,092 1,040 983

SCB Price inc. Cost Escalation - Nominal US$/t 130 137 141 144 140 136 125 115

SCB Price ex. Cost Escalation - Nominal US$/t 127 127 128 130 124 117 104 92

Consensus Relative Supply Position vs. 2012 Equilibrium Condition Mt 1,012 1,000 990 989 965 953 941 970

Consensus Price inc. Cost Escalation - Nominal US$/t 131 127 123 120 116 114 111 114

Consensus Price ex. Cost Escalation - Nominal US$/t 127 118 110 107 100 95 90 90

SCB vs. Consensus - Nominal US$/t 3 19 31 38 40 41 35 25

SCB vs. Consensus - Nominal % 2.7% 16.3% 28.1% 35.5% 40.2% 43.0% 39.2% 28.1%

SCB vs. Consensus - Due to Chinese Cost Escalation US$/t 4 9 13 14 16 19 21 24

SCB vs. Consensus - Due to S&D Differences US$/t () 10 18 24 24 22 14 2

SCB vs. Consensus - Due to Chinese Cost Escalation % 111.0% 49.3% 40.9% 36.8% 39.3% 45.4% 59.8% 93.3%

SCB vs. Consensus - Due to S&D Differences % -11.0% 50.7% 59.1% 63.2% 60.7% 54.6% 40.2% 6.7%


SO

N LA

PO

RT

E at P

ER

RY

CA

PIT

AL on 16-Jul-2013



Local Currency Costs and USD Revenues: The Stabilizing Effect of the "Natural Hedge" in Costs

As JK Galbraith observed, "faced with the choice between changing one's mind and proving that there is no need to do so, almost everyone gets busy on the proof." As an exercise in listening to Galbraith, we asked ourselves "what are the valuation impacts if we are wrong with regards to the iron ore price, and if $80/t iron ore becomes a near-term reality?" While examining this hypothetical situation, we came across three "automatic stabilizers" that would mitigate a part of the negative value impact of a sustained iron ore price fall for the big Western miners and offer a realistic picture of how a sustained fall in iron ore prices would impact valuations: (1) the natural FX hedge embedded in the miners' cost structures; (2) value-in-use (VIU) adjustments commanded by ores of differing qualities; and (3) elasticity of iron ore demand and spare capacity in China's steel-making industry whose exploitation would become economical, were the iron price to fall sufficiently. Across all of these stabilizers, the companies that are the most exposed to iron ore experience the greatest offsetting effect.

After several years of double-digit cost growth, the miners (particularly Rio Tinto and BHP Billiton) are winning the battle to contain further operating cost increases. Furthermore, in a world of falling commodity prices, we expect to see substantial weakening in producer currencies (AUD and BRL). Given that the miners' costs are denominated in local currency and their revenues in USD, any currency weakening will lead to an improvement in operating margin. Consequently, with nearly 50% of the cost increases in iron ore over the last 10 years coming from the impact of currency movements, any currency weakening is a significant source of upside. Our analysis shows that there is a natural hedge in the operating cost structure of the miners that softens the earnings impact of any sustained fall in commodity prices. In a world where iron ore falls to US$80/t, we estimate that Vale and Rio would experience the greatest benefits from this natural hedge, given their greater exposure to iron ore, while more diversified BHP and Anglo would experience lesser impacts.

FX as an Automatic Stabilizer The period from 2003 to 2012 saw the IMF 62% FE monthly spot (CFR Tianjin port) rise at a CAGR of 128%, from an average of US$14/t to an average of US$129/t. Over the same period, iron ore revenue and EBITDA for our coverage group rose at a CAGR of 133% and 141%, respectively. The combination of rising prices and volume increases has led to iron ore EBITDA margin expansion from, on average, 34% to 56%. While margin growth has been impressive, it could have been considerably more so, had the miners' costs not tracked prices upward.

Currencies in key iron ore producing countries show a strong correlation to iron ore prices. The R-squared for the Australian dollar and iron ore prices is 71% (see Exhibit 204). Similarly, for the Brazilian real, it is 57% (see Exhibit 205). The relationship in less iron-ore-intensive economies is weaker, e.g., the R-squared to the South African rand is only 15% (see Exhibit 206). It is therefore evident that iron ore price and cost rises are strongly linked through the medium of currency movement. Consequently, the R-squared between iron ore costs and revenue across our coverage group is 90% (see Exhibit 207). Fortunately, in presence of falling iron ore prices, the effect of this natural hedge is reversed.

In a World Where Iron Ore Falls to $80/t Iron Ore, We Estimate That Cost Savings — Combined With the Natural FX Hedge Embedded in Miners' Operating Cost Structure — Could Offset Valuation Declines by the Equivalent of, on Average, 6% of June 30, 2013 Share Prices for Rio, BHP, Vale and Anglo

The Strong Correlation Between Iron Ore Costs and Revenues Has Dampened Miners' Margin Expansion in a Rising Iron Ore Price Environment; If Iron Ore Prices Fall, the Effect of This Natural Hedge Is Reversed



AND LOVE THE ORE

As a consequence of mining being a commoditized business, there is a very strong natural hedge between its revenues and costs.38 Part of the relationship is driven by currency movements, which are influenced by balance of payment forces arising from strengthening commodity prices in producer countries. Specifically, mining companies based in a jurisdiction that uses a non-USD currency receive large amounts of USD as payments for the commodities they export. In order to compensate the labor and cover other mining costs in the mining jurisdiction, these companies have to convert the dollars they received into the local currency. In doing so, they increase both supply of dollars and demand for the local currency, thus inducing USD to depreciate vis-a-vis the local currency. Depreciation of the dollar then makes mining costs expressed in USD terms higher than they would be otherwise and reduces the miners' margins.

The second component of the natural hedge arises from the fact that significant elements in mining costs (like diesel or steel in grinding media or rubber for tires) are tied to the same macroeconomic drivers that influence commodity prices themselves. Looking at total iron ore costs, we find that local currency appreciation accounts for 42% of Rio Tinto’s iron ore costs and 51% of BHP's (see Exhibit 208 and Exhibit 209).

This also explains why hedging in mining is fraught with danger and, in our view, increases the level of operational risk rather than decreases it. If only the revenue is hedged, then the natural hedge between costs and revenue is broken. In addition, trying to hedge the costs of production as well as the revenue is extremely difficult. Moreover, there is always a price to be paid to set up any hedge, especially one as complicated as what we believe would be required to hedge all of the mining costs' components. There are numerous examples in mining where a naïve forward sale of material (often executed under the auspices of a risk management program) has destroyed significant value through breaking this natural hedge.

Exhibit 204 There Is a Strong Correlation Between the

Price of Iron Ore and the Strength of the Australian Dollar, With a High Iron Ore Price Equating to a Strong Local Currency

Exhibit 205 Likewise, the Relationship Exists (Albeit Less Pronounced) for the Brazilian Real


38 In the "Not All Iron Ores Were Created Equal: The Stabilizing Effect of Value in Use" chapter, we discuss how the degree of commoditization across different ores can vary, resulting in price differentials for different ores.

R² = 0.7147

0.70

0.80

0.90

1.00

1.10

1.20

1.30

1.40

1.50

1.60

1.70

0 25 50 75 100 125 150 175

1 U

SD

= X

AU

D

Iron Ore (US$/t)

Australian Dollar vs. Iron Ore Price

R² = 0.5718

1.00

1.50

2.00

2.50

3.00

3.50

0 25 50 75 100 125 150 175

1 U

SD

= X

BR

L

Iron Ore (US$/t)

Brazilian Real vs. Iron Ore Price




Exhibit 206 Unfortunately for Anglo American, for the South African Rand, the Relationship Is Not in Evidence

Exhibit 207 Mining Costs and Revenue Exhibit a High Degree of Correlation

Source: Bloomberg L.P. and Bernstein analysis. Source: Bloomberg L.P. and Bernstein analysis. Exhibit 208 Currency Appreciation Drives a Significant

Part of the Cost-to-Price Relationship…

Exhibit 209 ...Which Is Particularly True for the Australian Producers Rio Tinto and BHP Billiton


R² = 0.1516

5.00

5.50

6.00

6.50

7.00

7.50

8.00

8.50

9.00

9.50

0 25 50 75 100 125 150 175

1 U

SD

= X

ZA

R

Iron Ore (US$/t)

South African Rand vs. Iron Ore Price

R² = 0.9048

0

5,000

10,000

15,000

20,000

25,000

0 20,000 40,000 60,000

To

tal C

ove

rage

Co

sts

($m

)

Total Coverage Revenue

Total Iron Ore Costs and Revenue

41%

17%

42%

Rio Tinto Iron Ore Cost Drivers


Real Cost Increases

51%

22%

27%

BHP Iron Ore Cost Drivers


Real Cost Increases



AND LOVE THE ORE

Costs as an Automatic Stabilizer Macro effects notwithstanding, the miners have made significant progress in controlling iron ore costs, with Rio Tinto leading the way (see Exhibit 210 through Exhibit 212). It has shown a deceleration in the relationship between rising iron ore prices and costs as well as a slowdown in total cost growth. On top of that, Rio is the lowest-cost iron ore producer in our coverage. A component of this cost control is Rio Tinto's "Mine of the Future" program, which looks to increase the level of automation and remote automation in mining, thus reducing the requirement for ever more expensive labor. Rio Tinto believes that it represents a once-in-a-century step change in the way mining is conducted. Whether it will truly rate alongside the introduction of steam shovels or froth floatation remains to be seen, but it clearly has significant potential. In iron ore, for example, the program has a vision of utilizing remote-controlled drills to enable drill-and-blast extraction of ore; satellite-guided haul trucks to transport ore at the mine site; driverless trains to carry production to port; and teleoperated ship loading. Rio Tinto believes that the majority of the mining operations could be controlled from a center in Perth — some 1,500km away from the actual mine sites.

We applaud Rio Tinto's vision for the mining industry. Moreover, we look forward to seeing how this may continue to support Rio's ability to contain operating costs, and hence generate value creation.

Rio's fellow antipodean BHP Billiton has also achieved a marked deceleration in both cost growth over time and cost growth relative to iron ore price growth (see Exhibit 213 and Exhibit 214). This slowdown in cost growth coincided with a pause in a period of impressive growth for the company (during which cost controls can often be overlooked in favor of hitting volume targets). We believe that BHP's total cost profile is positively impacted by the nature of the company's most important iron ore asset — Mt Whaleback. This mine was established in 1968 and is the biggest single-pit, open-cut iron ore mine in the world. It is more than 5km long and nearly 1.5km wide. We believe that having large assets with significant inherent optionality greatly facilitates value accretive growth and enables miners to capture genuine economies of scale — trying to get growth from aggregations of numerous smaller assets will always be far more challenging.

Nevertheless, the company still has ground to cover. Its December half-year reporting noted that $1.9 billion in “controllable” cost savings (i.e., not fuel or taxation) had been achieved, and incoming CEO Andrew MacKenzie commented that he would "extend our pressure on costs and drive us to the bottom of cost per mined ton."39

Vale has continued to struggle with cost controls. Unlike its companions in the "Big Three," which have seen cost increases slow down, Vale’s costs have been growing at an accelerating rate (see Exhibit 215 and Exhibit 216). Interestingly, this has left Vale facing opex that is just 7% cheaper than BHP and 21% more expensive than Rio’s (all in USD terms). In its 1Q:13 earnings report, Vale noted that "for the first time in many years," costs and expenses were "an important source of improvement," with cost cutting extending beyond the operational into SG&A. The company reported savings of US$880 million year-over-year and US$2,539 million quarter-over-quarter (with US$1,283 million of this quarterly decline coming from COGS). While Vale’s average cost per mined ton of iron ore could certainly benefit from the new low-cost production from Serra Sul (cash opex estimated at ~US$30/t), we believe this to be an extraordinarily misguided lens through which to view cost controls. In our view, miners must consider the price impact that any new incremental tonnage has upon the market, as it impacts the price for all tons already in the market. For a functionally "pure" iron ore play like Vale, this is even more critical than for one of the more diversified miners.

39 http://www.smh.com.au/business/bhp-to-drive-focus-on-costs-in-new-era-20130220-2ermk.html.

The Miners Have Taken Steps (of Varying Efficacy) to Curb Iron Ore Cost Increases and Achieve Leaner Cost Structures Than in the Past




Anglo American has seen its iron ore costs rise though 2012 but showed some signs of containment (see Exhibit 217 and Exhibit 218). The story seems to be midway between the Australian and Brazilian members of the "Big Three," which is perhaps unsurprising, given the series of operational disappointments and the turnaround nature of Anglo's investment case. Perhaps equally unsurprising in this light is that if one Googles “Anglo American cost controls,” the first hits are not stories about CEOs trying to rein in costs in light of new projects (as with the Big Three); instead, they are for jobs as cost controllers within the company (see Exhibit 220).

Exhibit 210 Rio Tinto Is the Lowest Cost Iron Ore Producer in Our Coverage; Anglo, as the

Highest, Experiences Cash Opex That Is 50% Higher Than Rio's

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 211 While Mining Cost Escalation Is a Common

Worry, It Appears That the Miners Have Started to Get It Back Under Control

Exhibit 212 Rio Tinto, in Particular, Is Leading the Way



550

15

30

45

50

0 100

35

200 300 400 500 600 700 800150 250 350 450

10

5

50

0

25

40

650 750

20

Rio Vale BHP

33

49

4340

Anglo

+50%

Cumulative Production (2012 – Mt)

Opex–US$/t

R² = 0.9302

0

5

10

15

20

25

30

35

40

0 50 100 150 200

Un

it C

os

t (U

S$/

t)

Iron Ore Price — US$/t

Rio Tinto Iron Ore Costs vs. Price

0

5

10

15

20

25

30

35

40

Un

it C

os

ts (U

S$/

t)

Rio Tinto Iron Ore Costs Over Time



AND LOVE THE ORE

Exhibit 213 BHP Has Gone Through a Similar Experience With Both the Cost Relationship to Iron Ore...

Exhibit 214 ...And Over Time, Showing a Marked Deceleration



Exhibit 215 For Vale, the Evidence of Cost Control Is

Less Convincing...

Exhibit 216 …As Can Be Seen in the Continued Cost Growth Over Time



R² = 0.8923

0

5

10

15

20

25

30

35

40

45

50

0 50 100 150 200

Un

it C

os

t (U

S$/

t)


BHP Billiton Iron Ore Costs vs. Price

0

5

10

15

20

25

30

35

40

45

50

Un

it C

os

ts (U

S$/

t)

BHP Iron Ore Costs Over Time

R² = 0.7575

0

10

20

30

40

50

60

70

0 50 100 150 200

Un

it C

os

t (U

S$/

t)


Vale Iron Ore Costs vs. Price

0

10

20

30

40

50

60

70

Un

it C

os

ts (U

S$/

t)

Vale Iron Ore Costs Over Time




Exhibit 217 For Anglo, the Situation Seems to Lie Somewhere Between Vale and the Australians…

Exhibit 218 …With Some Evidence of Cost Containment



Exhibit 219 An Internet Search for the Miners and "Cost Controls" Results in an Ad for a Cost Control Position as at Least One of the Top 10 Hits (as of April 30, 2013); Only Anglo Has All of Its First Five Hits as Job Openings

Source: http://www.google.co.uk/search?q=anglo+american+cost+controls&rls=com.microsoft:en-gb&ie=UTF-8&oe=UTF-8&startIndex=&startPage=1&redir_esc=&ei=_zWAUdWkNMSCONaEgIgK.

R² = 0.8307

0

10

20

30

40

50

60

70

0 50 100 150 200

Un

it C

os

t (U

S$/

t)


Anglo Iron Ore Costs vs. Price

0

10

20

30

40

50

60

70

Un

it C

os

ts (U

S$/

t)

Anglo Iron Ore Costs Over Time



AND LOVE THE ORE

Impact for Our Coverage Declining iron ore prices would see producer currencies weaken, which would lower USD costs and improve margins for the miners. Based on the historical relationship between the iron ore price and iron ore producer currencies, and bearing in mind the current cost structures of our coverage companies (see Exhibit 211, Exhibit 213, Exhibit 215 and Exhibit 217), we estimate that the “natural hedge effect” linked to an iron ore price decline to US$80/t would offset valuation declines by 12% for Vale, 8% for Rio, 4% for BHP and have a neutral effect upon Anglo (all percentages relative to June 30, 2013 prices — see Exhibit 220 and Exhibit 221).40

Exhibit 220 Given the Historical Relationship Between the Iron Ore Price and Producer

Currencies, a Return to US$80/t Iron Ore Would See Producer Currencies Weaken, Lowering US$ Costs and Improving Margin for the Miners

Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 221 The Sensitivity to Such a Margin Impact Gives, on Average, ~6% Additional Upside

to the Australian and Brazilian Miners But Is Value Neutral for Anglo American


40 All valuation offsets are expressed as a percentage of the June 30, 2013 price, in USD terms.

12%

8%

4%

-1%Vale Rio BHP Anglo

% o

f Cur

rent

Sha

re P

rice

Share Price Impact of Curreny Effect of US$80/t Iron Ore

FX Impact Rio BHP Vale Anglo

Average $ Cost 32.7 42.5 40.0 49.1

Average LC Cost 32.8 42.6 70.7 374.7

US$ Cost at 80$ Fe 27.9 36.2 34.0 50.3

Margin Uplift per Ton ($) 4.9 6.3 6.0 -1.1

Tons Produced 2012 (kt) 246,831 161,149 303,443 49,137

Current EV/EBITDA 5.0 5.8 4.3 5.1

EV Impact ($m) 6,024 5,880 7,787 -283

Per Share Impact ($) 3.26 1.11 1.52 -0.22

June 30, 2013 Price ($) 40.38 25.46 12.90 18.40

Per Share Impact (%) 8% 4% 12% ‐1%

In a World of US$80/t Iron Ore, We Estimate That This Natural Hedge Could Offset Valuation Declines by, on Average, 6% of June 30, 2013 Share Prices




Not All Iron Ores Were Created Equal: The Stabilizing Effect of Value in Use

Not all iron ores were created equal. This chapter examines the second "automatic stabilizer" we have identified that would mitigate a part of the negative value impact of a sustained iron ore price fall for the big Western miners: Value-in-Use (VIU) and heterogeneity in iron ore products. The reference iron ore price is typically the 62% Fe landed China (CIF) price — the 62% grade is, of course, chosen to be broadly representative of much of the high-quality volume supplied into China from Australia and Brazil. However, a significant volume of Australian material (particularly for Fortescue Metals Group, not covered) suffers a substantial discount to the benchmark on the back of a lower Fe grade. Moreover, both Indian and Iranian material are important components of the overall supply mix into China (~10% of the total), and they also suffer a significant discount. This chapter examines the effective premium and discount on iron ores landed into China by region and by producer. We conclude that VIU offers a degree of automatic stabilization to the iron ore price, should a sustained period of "oversupply" begin to emerge. Furthermore, we expect the higher-quality producers to continue to produce in a declining price environment for longer. We believe that a not insubstantial portion of non-Chinese supply could become economically challenged and be pulled back from the market, should the reference price of iron ore experience sustained pressure. In a world where iron ore falls to US$80/t, we estimate that Vale and Rio would experience the greatest benefits from this natural hedge given their greater exposure to iron ore, while more diversified BHP and Anglo would experience lesser impacts.

Not All Iron Ores Were Created Equal In terms of geological abundance, iron ore is not scarce; it is the fourth most abundant element in the Earth's crust, comprising ~5% of the total. However, to be economically useful, the iron contained in iron ore must generally be 4-14x more concentrated (so 20% and 70% Fe by mass) than the average geological abundance (see Exhibit 222). While by no means rendering iron ore a scarce commodity, this does make a considerable difference to the types of material that can be exploited in the production of steel.

From a geological perspective, the story of economically viable iron ore begins in the Mesoarchean era (3,200 million to 2,800 million years ago), with the development of Banded Iron Formations (BIFs). The originally reducing environment of the very early Earth's atmosphere was progressively converted into an oxidizing (oxygen rich) atmosphere through the photosynthesis of blue-green algae in the Earth's oceans.41 The increasing levels of oxygen in the atmosphere reacted with the iron dissolved in the Earth's oceans to form insoluble iron oxides. As the levels of oxygen in seawater rose, the solubility of the iron oxides decreased. Following this decrease in solubility, the oxides were precipitated out of the seawater to form solid bands of iron-rich material. These bands were laid down intermingled with layers of mud and clay resulting in the characteristic pattern as shown in Exhibit 223. The iron-rich material is interwoven with silica-rich layers of

41An atmospheric condition rich in reducing gases (e.g., hydrogen and carbon monoxide) in which oxygen and other oxidizing gases and vapors are removed, thereby preventing oxidation.

VIU Offers a Degree of Automatic Stabilization to the Iron Ore Price, Should a Sustained Period of "Oversupply" Begin to Emerge; in a World Where Iron Ore Falls to US$80/t, This Stabilizer Could Offset Valuation Declines by, on Average, 7% of June 30, 2013 Share Price for Rio, BHP, Vale and Anglo

Iron Ore Is the Fourth Most Abundant Element, Comprising 5% of the Earth's Crust by Mass; However, Not All Iron Ores Are Created Equal



AND LOVE THE ORE

flint and chert42 to create BIFs. The second step in the generation of exploitable iron ore was the enrichment of this basic BIF. This occurred through the leaching out of the silica-rich layers, which led to concentration and further oxidation of the iron ores that were left behind. The weathering of the BIFs over millions of years effectively dissolved away the impurities to leave behind an enriched and economically valuable form of iron ore. It is this iron ore that is mined today. However, the process of weathering and enrichment of BIFs is complicated; the multiplicity of geological processes at work means that there are large variations in the material that can pass under the generic description of "iron ore."

A brief outline of some of the more important types of iron ore is given in Exhibit 224. The maximum Fe grade occurs at ~70%. This corresponds to chemically pure hematite or magnetite; at this degree of concentration, the residual mass is the oxygen associated with the iron rather than any impurity of the ore. The 62% referred to in the Fe benchmark is equivalent to 88.6% pure hematite (i.e., 62/70). It is the residual 11.4%, rather than 38%, that is non-iron bearing material.

As a further example of the point, we can take a single product — such as Vale's high-quality Standard Sinter Fines (SSF) — and look at what other than iron itself is actually in the iron ore (see Exhibit 225).

Exhibit 222 Rusting Granite in Peterborough, New Hampshire; During Pre-Industrial and Early Industrial Periods, Local Blacksmiths Relied Upon "Bog Iron" and "Mountain Iron" (Like the Below); While the Highest-Quality Deposit Reportedly Contained Ore of 56-63%, the Grade Was Often Much Lower and Was Quickly Depleted in Concentrations of Useful Economic Value

Note: A Gazetteer of the State of New-Hampshire (1823) by John Farmer and Jacob B. Moore noted that ore considered to be "the richest" in the nation was "obtained from a mountain in the east part of Concord [changed to Lisbon in 1824]." By 1849, however, The American Railroad Journal (vol. XXII, no. 609) noted that "The ores of New Hampshire, like those of Maine, are generally situated so that the expenses of transportation have rendered them of little value. Only one furnace is in operation, at Franconia…The furnace does but a small business, making only two and a half tons of iron a day." This stone furnace (no longer in use) is the only surviving iron smelter in New Hampshire today.

Source: Bernstein.

42 A fine-grained silica-rich sedimentary rock that often contains small fossils.




Exhibit 223 Where All the EBITDA Began — A Banded Iron Formation (BIF) Showing Characteristic Layering Patterns of Iron Ore and Gangue Material

Source: Wikimedia Commons. Exhibit 224 Not All Iron Ores Are Created Equal...

Source: Wikipedia and Bernstein analysis.

Ore Type Chemical Composition Max Fe

Magnetite Fe3O4 72.4%

Martite Fe3O4.Fe2O3 72.4% to 70.0%

Hematite Fe2O3 70.0%

Limonite 2Fe2O3.3H2O 59.8%

Goethite FeO.OH 62.9%

Siderite FeCO3 48.3%

Ilmenite FeO.TiO2 36.8%

Pyrite FeS2 46.7%



AND LOVE THE ORE

Exhibit 225 …And There Is a Significant "Tail" of Non-Iron Material in Iron Ores, All of Which Have a Value Implication

Source: IPCC and Bernstein analysis.

The purpose of this (very) brief geological digression was to make the point that not all iron ores were created equal. The 62% benchmark may generate a false impression of how homogenous — and hence how commoditized — iron ore is. The reality is that a variety of materials can be classified as "iron ore," all of which behave very differently in the blast furnace (BF). Consequently, each has a different VIU for a steel maker. The most important feature is that it is as costly to melt impurities as iron, and impurities yield no output benefit. Coke (the fuel source in BF steel making) is consumed to melt non-iron material lost to slag. Clearly, this is to the detriment of the steel maker.

Beyond wasting fuel, impurities also impact the resultant product. Many residual elements have a further deleterious effect on either the quality of the pig iron produced or the process of production. For example, phosphorous (P) renders steel brittle at low temperatures and so must be removed by a process of hot metal de-phosphorization, which incurs a cost. Alkali metals (K and Na) attack the refractory lining of the blast furnace, reducing campaign life as well as the strength of coke and pellets, and so lower overall productivity.

From an economic perspective, inclusion of a VIU adjustment into any cost curve tends to increase the marginal cost of production, thus putting upward pressure on the commodity's price. The removal of homogeneity in a commodity industry has, in some sense, a "de-commoditizing" effect — ceteris paribus, it increases the margin for advantaged players at the expense of those with disadvantaged ore assets (in the case of iron ore, a low Fe, high-contaminant profile). To the extent that these disadvantaged players are required to satisfy market demand, then the cost of the "benchmark" product must be high enough so that, once adjusted for VIU, it is still economic for their product to remain in the market.

Chemical Mass (% Wt)

Fe 64.50%

SiO2 4.50%

Al2O3 3.50%

Mn 0.150%

LOI (Loss on Ignition) 0.800%

P 0.027%

SiO2 0.007%

Na2O 0.015%

K2O 0.008%

TiO2 0.080%

CaO 0.020%

MgO 0.030%

From an Economic Perspective, Inclusion of a VIU Adjustment Into a Cost Curve Tends to Increase the Marginal Cost of Production, Thus Putting Upward Pressure on the Commodity's Price




VIU as an Automatic Stabilizer China is the world’s largest consumer of iron ore. In 2012, the Middle Kingdom accounted for 56% of global demand but produced just 13% of global supply. Australia and Brazil are the most important importers of iron ore to China (46% and 22% of 2012 imports, respectively, a proportion which has remained relatively constant over time) — see Exhibit 226 and Exhibit 227. The ore imported to satisfy this 42% gap between production and demand is of varying quality. Ore imported in March of this year, for example, ranges from 52% to 67% (see Exhibit 228). While this may not sound like a wide range, it is worth bearing in mind that the range is 3x the average concentration of iron ore in the Earth’s crust (5%). It is further worth considering that, as 70% is the maximum Fe grade, this range spans the equivalent of 74-96% pure hematite, with "impurities" ranging from 4% to 26%. These varying ores with their different grades and types of impurities, of course, command different premiums or discounts to the benchmark 62% Fe price (see Exhibit 229).

Unsurprisingly, iron ore grade accounts for 93% of the price variation (see Exhibit 230); the lowest grade imported (52%) commanded the greatest discount at 33%, while the highest grade imported (67%) commanded a 21% premium. However, even after adjusting for Fe grade, poorer-quality ores still command a significant discount (see Exhibit 231 through Exhibit 234). As a result, when the benchmark price of iron ore declines, exporters of poor-quality, low-grade iron ore into China can be expected to drop out faster than exporters of high-quality, high-grade iron ore. This functions like an automatic stabilizer, providing support not only to the iron ore price (thus slowing its rate of decline) but also to valuations of exporters from higher-quality deposits.

Exhibit 226 While Australia and Brazil Are the Most

Important Suppliers Into China, They Account for Only ~70% of the Total Non-Chinese Portion of Supply

Exhibit 227 The Overall Market Shares of Brazil and Australia Have Remained Roughly Constant Over Time; Perhaps This Indicates an Understanding from Incumbents That Competing on Volume in Commodity Markets Is Value Destructive


46%

22%

5%

4%

2%2%2%

2%

15%

China Iron Ore Imports by Origin

Aus Brazil SA India Iran

Canada Ukraine Russia Other

0%

10%

20%

30%

40%

50%

60%

Jan-

04

Aug

-04

Mar

-05

Oct

-05

May

-06

Dec

-06

Jul-0

7

Feb-

08

Sep

-08

Apr

-09

Nov

-09

Jun-

10

Jan-

11

Aug

-11

Mar

-12

Oct

-12

Share of Chinese Iron Ore Imports

Australia Share of Imports Brazil Share of Imports

The Price Paid for Iron Ore Imported Into China Varies Considerably; This Acts Like an Automatic Stabilizer in a Falling Price Environment, Providing Support Not Only to the Iron Ore Price But Also to Producers of High-Quality Product



AND LOVE THE ORE

Exhibit 228 There Is a Considerable Variation in the Observed Grade of Iron Ore Landed Into China — It Is Still Far from Being a Homogeneous Product

Source: BAIINFO and Bernstein analysis. Exhibit 229 An Analysis of 208 Pricing Points Shows a Significant Range of Variation Around the

Benchmark 62% CIF Price Point

Source: BAIINFO and Bernstein analysis.

50%

52%

54%

56%

58%

60%

62%

64%

66%

68%

Australia Brazil Chile India Iran Malaysia Mexico Russia South Africa

Ukraine

Fe

Gra

de

Distribution of Observed Iron Ore Grades Landed in China During March

R² = 0.7734

-40

-30

-20

-10

0

10

20

30

50% 52% 54% 56% 58% 60% 62% 64% 66% 68%

Iro

n O

re P

rem

ium

vs

. Be

nch

mar

k —

US

$/t

Fe Grade of Iron Ore

Price Impact of Quality in Iron Ore




Exhibit 230 Unsurprisingly, Fe Grade Explains Most of the Variation in Price

Exhibit 231 Nevertheless, Even After Adjusting for Fe Grade, Poor Ores Suffer a Significant Discount

Source: BAIINFO and Bernstein analysis. Source: BAIINFO and Bernstein analysis. Exhibit 232 This Is True Across Australian Iron Ores... Exhibit 233 ...As Well as Those from India

Source: BAIINFO and Bernstein analysis. Source: BAIINFO and Bernstein analysis.

R² = 0.9294

-40

-30

-20

-10

0

10

20

30

50% 55% 60% 65% 70%

Pre

miu

m —

US

$/t

Fe Grade

Average Premium (US$/t)

R² = 0.6298

-25

-20

-15

-10

-5

0

5

10

15

20

50% 55% 60% 65% 70%P

rem

ium

—U

S$/

tFe Grade

Average Premium (US¢/Dmtu)

R² = 0.9084

-20

-15

-10

-5

0

5

10

15

20

50% 55% 60% 65%

Pre

miu

m —

US

$/t

Fe Grade

Premium for Australian Ore

R² = 0.9938

-40

-35

-30

-25

-20

-15

-10

-5

0

5

50% 55% 60% 65%

Pre

miu

m —

US

$/t

Fe Grade

Premium for Indian Ore



AND LOVE THE ORE

Exhibit 234 There Is a US$46/t Differential Across Ores from Different Supply Locations

Source: BAIINFO and Bernstein estimates and analysis.

Impact for Our Coverage There is US$25/t of iron ore differential across the majors importing into China, with Fortescue Metals Group (not covered) at the low end of the range and Anglo American at the high end (see Exhibit 235). Assuming that 50% of the quality-disadvantaged material moves out of the market as prices begin to fall, VIU stabilizer can be expected to provide ~US$5/t of offsetting upward price support (see Exhibit 236). We estimate that Vale and Rio would experience the greatest benefits (equivalent to 11% and 9% of June 30, 2013 share prices), while more diversified BHP and Anglo experience 4% and 6% offsets, respectively.

Exhibit 235 Across the "Major" Space, There Is a US$25/t Differential

Source: BAIINFO and Bernstein estimates and analysis.

‐35

‐10

15

‐30

10

5

0

‐5

0 5 10 15 20 25 30 35 40 45 50

‐20

‐15

‐25

55

IranMalaysia

‐46

SAChile

Ukraine Brazil MexicoRussia Aus India

Average Monthly Imports to China ‐Mt

Prem

ium Paid in China

–US$/t

‐8

‐6

‐4

650‐2

2

4

550

6

600

8

12

14

350 400 450 500 700 750 800 850

‐12

0

10

0 50 100 150 200 250 300

‐10

BHPBRioVale FMG

‐25

Anglo

Annual Production by Majors ‐Mt

Prem

ium Paid in China

–US$/t

In a World of US$80/t Iron Ore, VIU Automatic Stabilizer Could Offset Valuation Declines by, on Average, 7% of June 30, 2013 Share Prices




Exhibit 236 Assuming That 50% of the Quality-Disadvantaged Material Moves Out of the Market as Price Begins to Fall, It Can Be Expected to Provide ~US$5/t of Offsetting Upward Price Support, Adding, on Average, $1.75/Share Across Our Coverage (as of June 30, 2013 Prices)


VIU Impact Rio BHP Vale Anglo

Average $ Cost 32.7 42.5 40.0 49.1

Margin Uplift per Ton ($) 5.4 5.4 5.4 5.4

Tons Produced 2012 (kt) 246,831 161,149 303,443 49,137


EV Impact ($m) 6,583 5,010 7,457 1,330

Per Share Impact ($) 3.56 0.94 1.46 1.04

June 30th, 2013 Price ($) 40.38 25.46 12.90 18.40

Per Share Impact (%) 9% 4% 11% 6%



AND LOVE THE ORE




If Prices Fall While Capital Stock Is Accumulating, Demand Accelerates: The Demand Elasticity Stabilizer

This chapter examines the third and final "automatic stabilizer" we have identified that would mitigate a part of the negative value impact of a sustained iron ore price fall for the big Western miners: the elasticity of demand for iron ore. Chinese steel making has seen its EBITDA margins cut by ~70% over the last decade on the back of rising raw material prices (iron ore CAGR 26%, PCI and coking coal CAGR 14%) and essentially stagnant steel prices (1% local currency CAGR, 4% USD CAGR). Commodity price escalation has driven steel making from being highly profitable to break even. We believe that this decline coincides with (or rather is instrumental in causing) the declines in Chinese steel production growth (down from ~20% p.a. 10 years ago to the ~6% p.a. seen today). However, steel capital stock in China is still only around one-third the level of that in the West, and the urbanization and industrialization of China remains far from complete. It is inconsistent to believe that China is undergoing a period of ongoing capital stock accumulation, while also believing that iron ore prices would fall as a consequence of low Chinese steel growth. Consequently, if iron ore prices were low for a sustained period of time on the back of an "oversupplied" market, demand would rise, which would itself unwind some of the apparent "oversupply." Compared to consensus expectations of a long-term iron ore price of 80US$/t, we believe that this feedback loop could generate long-term iron ore prices some 15US$/t higher than would otherwise be the case. This would see demand-supply equilibrium restored at a level closer to US$95/t. We estimate that Vale and Rio would experience the greatest benefits from this natural hedge, given their greater exposure to iron ore, while more diversified BHP and Anglo would experience lesser impacts.

China's Industrialization and Demographics — A Recap

According to our supply and demand overview discussed in the "Iron Ore 'Super-Cycle' Pricing Generated by High-Cost Producers Responding to the Rapidity of China's Industrialization" chapter, through months of data gathering, we assembled a database covering the consumption of metals and economic output for 120 countries since 1900. Our iron ore demand framework looks at two distinct components: the level of installed stock (here, cumulative steel intensity, i.e., total historical steel installed in an economy less depreciation) and the rate of steel consumption (here, current steel intensity). We look at both of these on a per-capita basis against GDP per capita, which enables us to account for not only the stage of industrialization (and output generation) of a given economy but also its importance to the overall global steel (and hence iron ore) consumption. Industrializing and industrialized economies move through three distinct phases of their rate of steel consumption (see Exhibit 237). The same three phases in the rate of steel consumption are repeated across industrialized and industrializing countries in our dataset. We categorize them as "Development," "Peak" and "Decline/Services." Furthermore, three paradigms emerge for the evolution of the level of steel consumption across countries that have industrialized (or attempted to). Two of these paradigms are successful ("Services and Technology," exemplified by the U.S. and the U.K., and "Manufacturing," exemplified by

If Iron Ore Prices Were Continuously Low on the Back of an "Oversupplied" Market, Demand Would Rise, Which Would Itself Unwind Some of the Apparent "Oversupply"; This Automatic Stabilizer Could Offset Valuation Declines by, on Average, 20% of the June 30, 2013 Share Prices for Vale, Rio, BHP and Anglo

Far from Being Overbuilt, China Resembles the U.S. in the 1930s, Has Just One-Third the per Capita Level of Embedded Steel as the Present Day U.S. and Is Facing Tremendous Pressure to Industrialize Before Its Old-Age Dependency Ratio Skyrockets



AND LOVE THE ORE

Germany and Japan), while one is not sustainable ("Inefficient," exemplified by the former USSR).

The 120 countries in our dataset display a strongly correlated relationship (R-squared = 66%) between levels of installed capital and GDP. Furthermore, and perhaps more importantly, as the outliers can be explained by individual economic circumstances (e.g., manufacturing hubs or endowment with oil), this framework generates a powerful predictive tool for looking at the sustainability of the current and future demand in the context of historical development. Our framework shows that Chinese steel demand (46% of 2012 finished steel demand globally) is on a sustainable path. China’s current steel intensity ("rate") is consistent with its stage of industrialization. It shows China transitioning from Phase 2 (peak) to Phase 3 (services) at a rate commensurate with that seen elsewhere. This is not a "structural shift" so much as a natural development to an economic state that is less steel intensive. To highlight the sustainability of China's demand, we would call attention to the fact that the economic activity of every person in the U.S. (or any other developed country) is supported by an installed capital base comprising 80kg of copper and 12,000kg of steel — well more than double the current capital stock in China (26kg of copper and 4,500kg of steel per capita).

Furthermore, as discussed in the "If China Is to Grow Rich Before It Grows Old, It Must Finish Industrializing in the Next Decade" chapter, a legacy of the "one child" policy is that China's aging population resembles an industrialized nation's demographics. However, it is supported by only a pre-industrial level of output. By 2020, China's working-age population will have peaked at close to 1 billion people. Over the next generation, 260 million people will leave the Chinese labor force, leading to a 90% increase in the old-age dependency ratio (see Exhibit 238). While such demographic challenges are not new, they have hitherto been associated with industrial rather than agrarian economies. By the time Japan's working population started to decline at the end of the 1990s, the country had an installed capital stock equivalent to nearly 14,000kg of steel per capita. With China's current capital stock standing at 4,500kg, our demand projections anticipate that this will approach 10,000kg by the time China's population starts to age dramatically (2020). In this regard, it is important to note that there is an important distinction between high young-age and old-age dependency ratios. Young-age dependency ratio = investment. I feed my children today on the

expectation that in the future they will feed not only themselves but will generate a surplus that will enable me to consume. The resources that are allocated toward a society's children are the quintessential investment in the future productive capacity of that society.

Old-age dependency ratio = consumption. Other than the social and moral imperative, it is not clear why I would want to feed granny. There is no hope that feeding granny will ever pay for itself. Allocating resources to the non-productive members of society can never generate a return in excess of the resources that are thereby consumed.

This is not an attempt to make any far-reaching social point. We emphasize that, from an economic perspective, one form of dependency ratio is "affordable" in the sense of offering a future return, while the other is not. If a society wishes to move to a gerontocracy, the imperative is to make sure that there is a level of productive capacity high enough to withstand the necessary intergenerational redistribution of the calls on that capacity without triggering social unrest.




Exhibit 237 Following the Same Three-Phase Trajectory for the Evolution of the Steel Consumption Rate Within All Industrialized and Industrializing Countries, China's Steel Intensity Has Passed from "Development" Through "Peak" and Is Now in "Decline"; the Steady State Rate of Consumption, Once the Process Is Completed, Will Be Determined by Whether the Economy, Once Mature, Adopts a More Manufacturing- or Services-Oriented Focus


0

10

20

30

40

50

60

70

80

0 5 10 15 20 25 30 35 40 45 50

Ste

el In

ten

sity

(kg

/$'0

00

GD

P)






AND LOVE THE ORE

Exhibit 238 In 2000, China Represented a Clear Outlier from the Normal Pattern: the World's Largest Country Was Set to Age Rapidly and Yet Was Still at a Pre-Industrial Level of Output; by 2020, China's Working-Age Population Will Have Peaked; Over the Next Generation, 260 Million People Will Leave the Chinese Labor Force, Leading to a 90% Increase in the Old-Age Dependency Ratio


-10

0

10

20

30

40

50

60

-50 -40 -30 -20 -10 0 10

GD

P/C

apit

a —

US

$'00

0



USA

Japan

China

South Korea




Lower Iron Ore Price as a Stimulus for Chinese Steel Making Over the last decade, the Chinese domestic price of steel has grown at a CAGR of 1% (4% in USD terms), while its input costs have risen at, on average, 8.5%. In USD terms, iron ore costs have had a CAGR of 26% over the period, while for both PCI and coking coal the CAGR has been 14% (see Exhibit 239 through Exhibit 243). As a result, raw material input costs per ton of steel have risen from less than US$100/t in 2003 to more than US$300/t today (see Exhibit 244) and comprise nearly three-fourths of the overall costs of steel making (see Exhibit 245 and Exhibit 246). The rising input costs, combined with more slowly rising steel prices over the last decade, have, unsurprisingly, seen steel-making EBITDA margins cut to one-third of their previous levels — from an average of 62% in 2003 to 21% by the end of 1Q:13 (see Exhibit 247) — and profit margins all but wiped out.

While at first glance Chinese crude steel production appears to have grown steadily (see Exhibit 248), this belies the fact that steel production growth has been declining. Concurrent with the rising input cost pressures, the annualized crude steel production growth has declined from ~20% to ~6% (see Exhibit 249 and Exhibit 250). The correlation between steel growth and steel profitability has an R-squared of 52% (see Exhibit 251).

Exhibit 239 Over the Last 10 Years, the Domestic Steel

Price in China Has Been Remarkably Constant...

Exhibit 240 ...Whereas the Price of Raw Materials Has Accelerated Sharply; It Is True for All — Iron Ore...

Source: Bloomberg L.P. and Bernstein analysis. Source: Bloomberg L.P., IMF and Bernstein analysis.

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

Jan-

03

Sep

-03

May

-…

Jan-

05

Sep

-05

May

-…

Jan-

07

Sep

-07

May

-…

Jan-

09

Sep

-09

May

-…

Jan-

11

Sep

-11

May

-…

Jan-

13

HR

C —

RM

B/t

Chinese Domestic Steel Price

0

20

40

60

80

100

120

140

160

180

200Ja

n-0

3

Sep

-03

May

-04

Jan

-05

Sep

-05

Ma y

-06

Jan

-07

Se p

-07

May

-08

Jan

-09

Se p

-09

May

-10

Jan

-11

Sep

-11

Ma y

-12

Jan

-13

US

$/t

Iron Ore Price

In the West, a Negative Step Change in Steel Demand Growth Occurred Only Once Urbanization Was Complete; We Believe That If the Price of Iron Ore Were to Fall, the Demand for Steel Would Rise



AND LOVE THE ORE

Exhibit 241 ...Key Fuel and Reductants Used in Steel Making — HCC...

Exhibit 242 ...And Its Replacement PCI

Source: CRU and Bernstein analysis. Source: CRU and Bernstein analysis. Exhibit 243 The Rise in Iron Ore Price Has Done the Most to Increase the Costs of Chinese Steel

Production, Reflecting a Transfer of Value from the Chinese Industrial Sector to the Western-Owned Miners

Source: AME, CRU, IMF and Bernstein analysis.

0

50

100

150

200

250

300

350Ja

n-0

3A

ug-0

3M

ar-0

4O

ct-0

4M

ay-0

5D

ec-0

5Ju

l-06

Feb

-07

Sep

-07

Ap

r-08

No

v-08

Jun

-09

Jan

-10

Aug

-10

Mar

-11

Oct

-11

May

-12

Dec

-12

US

$/t

Hard Coking Coal (HCC) Price

0

50

100

150

200

250

300

Jan

-03

Sep

-03

May

-04

Jan

-05

Sep

-05

Ma y

-06

Jan

-07

Sep

-07

Ma y

-08

Jan

-09

Se p

-09

Ma y

-10

Jan

-11

Sep

-11

Ma y

-12

Jan

-13

US

$/t

PCI Price

26%

14% 14%

6%5% 5%

4% 4% 3% 3%

Iron Ore Cost

PCI Cost Coking Coal Cost

Steel Making

Iron Making Sintering Admin & Support

Rolling Casting & Refining

Coke Making

CA

GR

Chinese Steel-Making Cost Escalation (CAGR 2003-2013)




Exhibit 244 Raw Materials Fed Into Steel Making Used to Cost Less Than 100US$/t; It Is Now Well Over 300US$/t

Source: AME, CRU, IMF and Bernstein analysis. Exhibit 245 Currently, the Raw Material Cost of Steel

Making Accounts for Nearly Three Quarters of the Overall Costs in China

Exhibit 246 There Has Also Been a Slower, But Nonetheless Pronounced Rise in the Non-Raw Materials Costs of Steel Making in China

Source: AME, CRU, IMF and Bernstein analysis. Source: AME and Bernstein analysis.

0

100

200

300

400

500

600Ja

n-03

May

-03

Sep

-03

Jan-

04

May

-04

Sep

-04

Jan-

05

May

-05

Sep

-05

Jan-

06

May

-06

Sep

-06

Jan-

07

May

-07

Sep

-07

Jan-

08

Ma y

-08

Sep

-08

Jan-

09

May

-09

Sep

-09

Jan-

10

Ma y

-10

Sep

-10

Jan-

11

Ma y

-11

Se p

-11

Jan-

12

May

-12

Sep

-12

Jan-

13

US

$/t

Chinese Steel-Making Raw Materials Cost

Iron Ore Cost Coking Coal Cost PCI Cost

45%

21%

3%

31%

Steel Making Costs by Element (2013)

Iron Ore Cost Coking Coal Cost PCI Cost Other Costs

5540

35

24

27

14

15

9

11

7

7

4

2

2

2013 2003

US

$/t S

tee

l

Non-Raw Material Cost of Steel Making

Casting & Refining Rolling Steel MakingIron Making Admin & Support SinteringCoke Making



AND LOVE THE ORE

Exhibit 247 Unsurprisingly, Margins in the Chinese Steel-Making Industry Have Declined on the Back of Higher Raw Material Prices

Source: AME, CRU, IMF and Bernstein analysis. Exhibit 248 While, at First Glance, the Growth in Chinese Steel Production Appears Monotonic...

Source: Bloomberg L.P., AME, CRU, IMF and Bernstein analysis.

0%

10%

20%

30%

40%

50%

60%

70%

80%Ja

n-03

May

-03

Se p

-03

Jan-

04

May

-04

Sep

-04

Jan-

05

Ma y

-05

Sep

-05

Jan-

06

Ma y

-06

Sep

-06

Jan-

07

Ma y

-07

Sep

-07

Jan-

08

Ma y

-08

Se p

-08

Jan-

09

May

-09

Se p

-09

Jan-

10

Ma y

-10

Se p

-10

Jan-

11

May

-11

Se p

-11

Jan-

12

May

-12

Se p

-12

Jan-

13

HR

C E

BIT

DA

Mar

gin

Chinese Steel-Making Margin

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

Jan-

03

May

-03

Se p

-03

Jan-

04

May

-04

Sep

-04

Jan-

05

May

-05

Se p

-05

Jan-

06

Ma y

-06

Sep

-06

Jan-

07

May

-07

Sep

-07

Jan-

08

May

-08

Sep

-08

Jan-

09

May

-09

Sep

-09

Jan-

10

May

-10

Se p

-10

Jan-

11

May

-11

Sep

-11

Jan-

12

Ma y

-12

Se p

-12

Jan-

13

Cru

de

Ste

el —

kt

Chinese Monthly Crude Steel Production




Exhibit 249 ...This Belies a Marked Slowdown in Chinese Steel Production Rates from ~20% at the Start of the Last Decade to Closer to 6% Today, Concurrent With the Declining Steel-Making Margins

Source: Bloomberg L.P. and Bernstein analysis. Exhibit 250 Growth and Profitability Declined Concurrently

Source: Bloomberg L.P., AME, CRU, IMF and Bernstein analysis.

-10%

-5%

0%

5%

10%

15%

20%

25%

30%

35%Ja

n-05

Apr

-05

Jul-0

5

Oct

-05

Jan-

06

Apr

-06

Jul-0

6

Oct

-06

Jan-

07

Apr

-07

Jul-0

7

Oct

-07

Jan-

08

Apr

-08

Jul-0

8

Oct

-08

Jan-

09

Apr

-09

Jul-0

9

Oct

-09

Jan-

10

Apr

-10

Jul-1

0

Oct

-10

Jan-

11

Apr

-11

Jul-1

1

Oct

-11

Jan-

12

Apr

-12

Jul-1

2

Oct

-12

Jan-

13

An

nu

aliz

ed T

wo-

Yea

r MoM

CA

GR

Crude Steel Production Growth Rate

-10%

-5%

0%

5%

10%

15%

20%

25%

30%

35%

0%

10%

20%

30%

40%

50%

60%

70%

80%

Jan-

05

Apr

-05

Jul-0

5

Oct

-05

Jan-

06

Apr

-06

Jul-0

6

Oct

-06

Jan-

07

Apr

-07

Jul-0

7

Oct

-07

Jan-

08

Apr

-08

Jul-0

8

Oct

-08

Jan-

09

Apr

-09

Jul-0

9

Oct

-09

Jan-

10

Apr

-10

Jul-1

0

Oct

-10

Jan-

11

Apr

-11

Jul-1

1

Oct

-11

Jan-

12

Apr

-12

Jul-1

2

Oct

-12

Jan-

13

Ste

el G

row

th

Ste

el E

BIT

DA

Mar

gin


Margin, % Annualized Grow th



AND LOVE THE ORE

Exhibit 251 We Believe That a Significant Part of the Slowdown in Chinese Steel Growth Resulted from High Commodity Prices; the West Experienced a Step Change in Steel Demand Growth Only After Urbanization Was Complete; However, This Is Far from Complete in China, Combined With the Imminent Demographic Shift Enables Us to Conclude That in the Presence of Falling Iron Price, Steel Demand Would Rise


While it is clear that correlation does not necessarily imply causation, we believe that a significant part of the slowdown in Chinese steel growth was a result of the budgetary constraint imposed by high commodity prices. In the West, a negative step change in steel demand growth occurred only once urbanization was complete. Urbanization is far from complete in China. Consequently, we believe that if the price of iron ore were to fall, the demand for steel would rise (particularly bearing in mind the hard stop the demographic shift in China is imposing upon the industrialization process). Consequently, we believe that it is incorrect to interpret the step change seen in OECD steel demand growth post the 1970s' oil shocks as representative of the situation in China today. By the time the OECD-trend steel demand declined, urbanization was complete, and this is not the case in China today. Rather, when China "unleashed" the forces of private enterprise at the turn of the century, it enabled the industry to actively chase the enormous profitability that was available as a consequence of combining the productivity of large blast furnaces with the vast pools of surplus Chinese labor and the availability of incredibly cheap iron ore on the international commodity markets. As commodity prices re-equilibrated themselves, this excess return was passed from Chinese steel makers on to Australian (and Brazilian) iron ore producers. The effect of this is, naturally enough, to reduce the incentive to grow steel output, while increasing the incentive to grow iron ore output. To the extent that the incumbent iron ore producers wish to unwind the structural advantage they currently possess through continuing the poorly thoughtout strategy of value-destructive capital profligacy, they have at least this one consolation: lowering the price of iron ore ought to encourage a higher level of Chinese demand.

A complicating factor in the analysis of steel elasticity is the effect of Chinese monetary stimulus of 2009. A strong relationship exists between lagged steel growth and the growth in the money supply (R-squared = 55%) — see Exhibit 252 through Exhibit 254. If one strips out the effect of Chinese monetary stimulus in 2009, the correlation between steel growth and EBITDA margin increases (R-

R² = 0.5187

-5%

0%

5%

10%

15%

20%

25%

30%

35%

0% 10% 20% 30% 40% 50% 60% 70% 80%

An

nu

aliz

ed S

teel

Gro

wth

Rat

e

Chinese Steel EBITDA Margin

Chinese Steel Growth vs. Chinese Steel Profitability




squared = 67%) — see Exhibit 255 and Exhibit 256. Stripping out the effect of M2 growth suppresses the steel growth rate that one would associate with any given industry profitability. Consequently, as well as increasing the overall correlation, stripping out government stimulus embeds a level of conservatism in the forward-looking counterfactual demand growth that we calculate for the falling iron ore price scenario. A further level of conservatism is factored in by allowing for a declining steel price in our analysis (down to an HRC price of US$600/t), which reflects the expectation of continuing steel price declines on the back of sustained "oversupply" in steel. On the other hand, were the steel price to stay higher than this, the effect would be to further improve the profitability (and growth) of the Chinese steel industry.

Exhibit 252 A Complicating Factor in the Analysis Steel

Elasticity Is the Effect of Chinese Monetary Stimulus in 2009

Exhibit 253 There Is a Strong Relationship Between Lagged Steel Growth and the Growth in the Supply of Money...

Source: Bloomberg L.P. and Bernstein analysis. Source: Bloomberg L.P. and Bernstein analysis. Exhibit 254 ...We Strip This Effect Out of the Analysis of Steel Demand Growth...


0%

5%

10%

15%

20%

25%

30%

35%

Jan

-07

Jun

-07

No

v-07

Ap

r-08

Sep

-08

Feb

-09

Jul-

09

Dec

-09

May

-10

Oct

-10

Mar

-11

Aug

-11

Jan

-12

Jun

-12

No

v-12

M2

Yo

Y S

up

ply

Gro

wth

Chinese Money Supply

0%

5%

10%

15%

20%

25%

0%

5%

10%

15%

20%

25%

30%

35%

Jan

-09

Ap

r-09

Jul-

09

Oct

-09

Jan

-10

Ap

r-10

Jul-

10

Oct

-10

Jan

-11

Ap

r-11

Jul-

11

Oct

-11

Ste

el

M2

M2 and Lagged Steel

M2 Growth Steel Growth

R² = 0.5537

0%

5%

10%

15%

20%

25%

0% 10% 20% 30% 40%

Lag

ged

Ste

el G

row

th

M2 Supply Growth

Chinese Money Supply vs. Steel Growth



AND LOVE THE ORE

Exhibit 255 ...The Effect of This Would Be to Lower the Level of Steel Production for Any Given Profitability...

Source: Bloomberg L.P., AME, CRU, IMF and Bernstein estimates and analysis. Exhibit 256 ...Which Increases the Strength of the Relationship


-10%

-5%

0%

5%

10%

15%

20%

25%

30%

35%

0%

10%

20%

30%

40%

50%

60%

70%

80%Ja

n-05

Apr

-05

Jul-0

5

Oct

-05

Jan-

06

Apr

-06

Jul-0

6

Oct

-06

Jan-

07

Apr

-07

Jul-0

7

Oct

-07

Jan-

08

Apr

-08

Jul-0

8

Oct

-08

Jan-

09

Apr

-09

Jul-0

9

Oct

-09

Jan-

10

Apr

-10

Jul-1

0

Oct

-10

Jan-

11

Apr

-11

Jul-1

1

Oct

-11

Jan-

12

Apr

-12

Jul-1

2

Oct

-12

Jan-

13

Ste

el G

row

th

Ste

el E

BIT

DA

Mar

gin


Margin, % Annualized Grow th Steel Grow th Ex-Monetary Grow th

R² = 0.6679

-5%

0%

5%

10%

15%

20%

25%

30%

35%

0% 10% 20% 30% 40% 50% 60% 70% 80%

Ste

el G

row

th

Steel EBITDA Margin

Chinese Steel Growth Ex-Impact of M2 Growth




Impact on Our Coverage Valuation Accepting that the Chinese government is industrializing China as fast as it can, thanks to demographic pressures (and from the equivalent of a capital stock base akin to that of the U.S. in the 1930s), and seeing the margin compression that has occurred in the steel industry, as iron ore prices (and other inputs) have risen over the last decade, mean accepting that falling iron ore prices would lead to increased steel growth. In a world where iron ore falls to $80/t, we estimate this reduced cost pressure would increase steel trend growth rates by 3% — an increase in demand, which, we estimate, would in fact push iron ore price back up to $95/t (see Exhibit 257 and Exhibit 258).

Companies with the highest iron ore exposure would of course benefit the most; we estimate the offset for Vale would be the equivalent of 31% of its June 30, 2013 share price, Rio 24%, Anglo 16%, and BHP 10% (see Exhibit 259).

Exhibit 257 It Is Inconsistent to Simultaneously Believe That Iron Ore Prices Would Fall as a

Consequence of Low Chinese Steel Growth and That China Is Undergoing a Period of Ongoing Capital Stock Accumulation; Consequently, We Believe That Falling Iron Ore Prices Would Lead to Increased Steel Growth of an Additional 3% Impact on Trend Steel Growth Rates, Even Given a Declining Steel Price


Average 2011 to Present Forecast at 80US$/t Fe

Steel Price - $/t 673 600

Iron Ore Price - $/t 148 80

Iron Ore Cost - $/t 237 128

Fuel Costs - $/t 172 172

Other Costs - $/t 152 152

Total Costs - $/t 561 452

Margin - $/t 112 148

Margin - % 16.7% 24.7%

Best Fit Steel Growth Rate - % 8.1% 10.9%

Observed Steel Growth Rate - % 8.0% N/A

Impact of Stimulus - % -1.9% -1.9%

Trend Growth Rate - % 6.3% 9.0%

Impact of Lower Iron Ore Prices - % 2.7%

Compared to Consensus Expectations of a Long-Term Iron Ore Price of US$80/t (Which Is Already Being Discounted by Mining Equities), We Estimate That This Automatic Stabilizer Could Provide an Additional US$15/t of Support Leading to an Equilibrium Demand-Supply Situation Closer to US$95/t



AND LOVE THE ORE

Exhibit 258 The Impact of the Additional Steel Demand Helping Offset the Impact of "Oversupply" from New Projects Could Support a Higher Price Level Than Would Otherwise Be Achieved

Source: Bloomberg L.P. and Bernstein estimates and analysis. Exhibit 259 The Impact of Incorporating This Effect Could See Iron Ore Prices Stabilize Some

15US$/t Higher Than Would Be the Case If the Elasticity of Demand Is Ignored


0

20

40

60

80

100

120

140

1600 50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

850

900

950

1,00

0

1,05

0

1,10

0

1,15

0

1,20

0

1,25

0

1,30

0

1,35

0

1,40

0

1,45

0

1,50

0

1,55

0

1,60

0

Iro

n O

re (U

S$/

t)

Current and Future Iron Ore Cost Curves

Current Cost Curve Future Cost Curve

Price equillibrium higher as a conseuence of incorporating

demand elasticity

Price equilibrium higher as a conseuence of incorporating

demand elasticity

Lowering priceincreases trend

demand

Elasticity Impact Rio BHP Vale Anglo

Average $ Cost 32.7 42.5 40.0 49.1

Margin Uplift per Ton ($) 14.9 14.9 14.9 14.9

Tons Produced 2012 (kt) 246,831 161,149 303,443 49,137


EV Impact ($m) 18,255 13,893 20,680 3,689

Per Share Impact ($) 9.89 2.61 4.04 2.89

June 30, 2013 Price ($) 40.38 25.46 12.90 18.40

Per Share Impact (%) 24% 10% 31% 16%




Valuation and Risks

We provide two valuation metrics: As mining companies are operationally and financially geared to their underlying

commodity baskets (~80% of weekly equity price moves can be explained by moves in the underlying commodity prices), we use a regression-based trading model and our forward-looking commodity price forecasts to help determine our 12-month target prices (and to generate six monthly price forecasts). To the extent that the regression holds, and the parameters of the regression have not significantly shifted, we take the target price from the trading model. To the extent that the regression is shifting or the equity is deviating, we look for evidence of whether this shift or deviation is temporary (and hence may be expected to close) or whether it signals a more fundamental re- or de-rating of the equity. In the event that there is no significant deviation or if we believe a deviation is temporary, the target price is set by the trading model. In the event that we believe a deviation is signaling a fundamental change, we will adjust our target price for this fundamental shift and disclose the manner and magnitude of the adjustment made. At present, no adjustments have been made to the target prices generated by our trading model. Note that we round final target prices in 25p/cent increments. Exhibit 260 summarizes our ratings and target prices as of June 30, 2013.

In addition to the target price (and short-term price forecasts generated by our trading model), we provide a supplementary valuation based on DCF. Given the long-lived nature of mining assets, we believe a DCF is critical to understanding the intrinsic value of a share (what the share price, in our view, "ought" to be today). Our DCF model constructed in nominal local currency terms out to 2030, over which explicit commodity price and exchange rate forecasts apply. The nominal local currency cash flows are de-escalated into real U.S. dollar cash flows and discounted at the company-specific WACC. A country risk premium reflecting the geographic origin of the cash flows is added to the underlying WACC to reflect cash-flow items (i.e., expropriation) that cannot be explicitly modeled in the cash flow. All reserves are considered exploited by the model. In addition, 50% of the incremental resources (i.e., 50% of the residual resources excluding those that have already been converted to reserves) of the company are modeled. Where residual life of mine (LOM) may be inferred for operations beyond the 2030 time horizon, a terminal value is calculated for the remaining years of potentially exploitable material. We use this methodology to derive all forward-looking multiples and other valuation metrics. Note that we forecast our models in reporting currency (USD), convert to listing currency (GBP or Real) at an average exchange rate, and round final DCF values in 25p/cent increments.

Investing in the Miners Is Always a Call on the Future Prices of Their Commodity Exposure



AND LOVE THE ORE

The Trading Model Sets 12-Month Target Prices The value of the miners is essentially a call on the future commodity price deck; each of the miners shows a very strong relationship to their underlying commodity baskets. Our regression-based trading model looks at the relationship between a company's historical share price performance and its underlying commodity exposure. It fits the best estimate relationship between the mining company valuation and its commodity basket historical performance, and then rolls that relationship forward over time. We do this to capture the fact that the parameterization of the fit between the miners and the commodity is not fixed.

Regression-based trading models must be taken, of course, with a grain of salt. A strong regression or R-squared fit between two variables can of course be preserved, while changing the parameterization of that fit. In other words, that a relationship exists may always be true, but the nature of that relationship may change over time. It is the changing nature of the relationship between a mining equity and its underlying commodity basket that expresses the de-rating or re-rating of a sector.

Accordingly, we monitor how the results generated by this model change over time. Does the best-fit change (and if so, for one stock or multiple stocks)? And is the equity deviating from its historical relationship (or is the R-squared changing over time)? To the extent that the regression is shifting or the equity is deviating, we look for evidence of whether this shift or deviation is temporary (and hence may be expected to close) or whether it signals a more fundamental re- or de-rating of the equity. In the event that there is no significant deviation or if we believe a deviation is temporary, the target price is set by the trading model. In the event that we believe a deviation is signaling a fundamental change, we will adjust our target price for this fundamental shift and disclose the precise adjustment made. At present, no adjustments have been made to the target prices generated by our trading model. Exhibit 261 through Exhibit 280 summarize our ratings and target prices as of June 30, 2013.43

43 In order to generate a regression analysis for the new Glencore Xstrata entity, we calculate a synthetic price for the entity based on the trading

history of Xstrata and the relationship that existed between Glencore and Xstrata post the Glencore listing but before the Xstrata offer. Clearly, this is a less reliable basis than for the other miners but we believe that it continues to have some validity.

Our Trading Model Starts With Regression Analysis — But We Caution That One Must Interpret the Regressions Before Setting Target Prices




Exhibit 260 Summary of Our Coverage

Source: Bloomberg L.P., FactSet and Bernstein estimates and analysis.

BHP’s regression has been tight, not only in our most recent model update conducted June 5, 2013, but in our previous trading model updates. The R-squared of the most recent update is 78% (see Exhibit 261 and Exhibit 262). At the time of our update, BHP was moderately overvalued relative to its best-fit regression (see Exhibit 263). We find this unsurprising given the faltering market sentiment and the perception (justified in our view) of BHP as the highest quality stock in our coverage. Our trading model generates a 12-month target price of £22.50 (see Exhibit 264). We have not seen signs of the regression shifting nor is there anything in our fundamental research to date that causes us to anticipate it will; hence, we set our target price from the unmodified regression model price.

7/12/2013 Anglo BHPB Rio Tinto Vale

£ £ £ BRL

Price 12.95 18.00 27.99 30.32

Price Target 20.25 22.50 41.25 46.50

Potential Up/Downside 56% 25% 47% 53%

EV/EBITDA ‐ Current 5.09 5.67 4.89 4.30

EV/EBITDA ‐ 5 Yr. Avg. 4.9 5.5 4.8 5.0

EV/EBITDA Target 8.4 8.6 9.8 8.2

PE ‐ Current 8.9 9.3 7.1 5.5

PE ‐ 5 Yr. Avg. 9.2 9.4 8.0 7.6

PE Target 16.1 14.0 16.8 21.7

2013 EPS ‐ Consensus 2.11 2.47 5.65 2.48

2013 EPS ‐ SCB 1.90 2.70 5.33 2.28

% SCB vs. Consensus (9.8%) 9.4% (5.6%) (8.2%)

2014 EPS ‐ Consensus 2.63 2.84 6.44 2.89

2014 EPS ‐ SCB 2.57 3.73 7.34 2.55

% SCB vs. Consensus (2.2%) 31.3% 14.0% (11.8%)

*Note all EPS figures are in reporting currency (USD)

BHP: TP £22.50 (Unmodified from Trading Model Prediction)



AND LOVE THE ORE

Exhibit 261 The Relationship Between the Miners and Their Underlying Commodity Exposure Is Incredibly Strong...

Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 262 ...And Explains the Majority of the Historical Performance


50,000

70,000

90,000

110,000

130,000

150,000

170,000

190,000

210,000

230,000

250,000

Jun/06 Dec/06 Jun/07 Dec/07 Jun/08 Dec/08 Jun/09 Dec/09 Jun/10 Dec/10 Jun/11 Dec/11 Jun/12 Dec/12

BH

P E

V (U

S$m

)

BHP EV — Fitted vs. Actual

Fitted Actual

R² = 0.7848

50,000

70,000

90,000

110,000

130,000

150,000

170,000

190,000

210,000

230,000

250,000

50,000 70,000 90,000 110,000 130,000 150,000 170,000 190,000 210,000 230,000 250,000

BH

P A

ctu

al E

V (

US

$m)

BHP Fitted EV (US$m)

BHP Scatter Plot — Fitted vs. Actual




Exhibit 263 BHP Is Currently Slightly Overvalued Relative to Its Historical Best Fit, an Unsurprising Development in the Current Tremulous Market Environment

Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 264 We See Considerable Upside in the Share Price


Like BHP, Rio Tinto’s regression has also been tight in both our historical and most recent trading model update (June 5, 2013). The R-squared of the most recent update is 88% (see Exhibit 265 and Exhibit 266). At the time of our update, Rio was slightly overvalued relative to its best-fit regression (see Exhibit 267), but by less than 5%. We consider this unsurprising, given the recent slides in both iron ore prices and Rio Tinto’s share price (bearing in mind that iron ore generated ~90% of Rio Tinto’s 2012 EBIT). Our trading model generates a 12-month target price of £41.25 (see Exhibit 268). We have not seen signs of the regression shifting nor is

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%


% U

nd

er/O

ve

r V

alu

ati

on

BHP Share Price vs. Best Fit Regression

Undervalued Relative to Commodity Basket

Overvalued Relative to Commodity Basket

0

5

10

15

20

25

30

35

40

Sh

are

Pri

ce

(£)

BHPB Share Price Forecast

Actual Forecast ex Volume Growth Forecast Inc Volume Growth DCF SOTP

Rio Tinto: TP £41.25 (Unmodified from Trading Model Prediction)



AND LOVE THE ORE

there anything in our fundamental research to date that causes us to anticipate it will; hence, we set our target price from the unmodified regression model price.

Exhibit 265 The Relationship Between the Miners and Their Underlying Commodity Exposure Is

Incredibly Strong…

Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 266 …And Explains the Majority of the Historical Performance


50,000

70,000

90,000

110,000

130,000

150,000

170,000

190,000

210,000

230,000

250,000

Jun-06 Dec-06 Jun-07 Dec-07 Jun-08 Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Jun-11 Dec-11 Jun-12 Dec-12

Rio

EV

(US

$m)

Rio EV — Fitted vs. Actual

Fitted Actual

R² = 0.887

50,000

70,000

90,000

110,000

130,000

150,000

170,000

190,000

210,000

230,000

250,000

50,000 70,000 90,000 110,000 130,000 150,000 170,000 190,000 210,000 230,000 250,000

Rio

Act

ual

EV

(U

S$m

)


Rio Scatter Plot — Fitted vs. Actual




Exhibit 267 Rio Is Currently Slightly Overvalued Relative to Its Historical Best Fit, But Less Than 5%, and Hence Less Than High-Quality BHP



In our previous trading model update, we noted that Anglo American’s equity price had declined from fairly valued at the start of the year relative to its commodity basket to 12% below the basket’s spot price (see European Metals & Mining: Timing Is Everything. Predictions From Our Short-Term Trading Model).

At the time we commented that whether the deviation was temporary or resulted in a permanent de-rating would depend, in our view, upon the resolution of the company's operating disappointments and South African risk. With our June 5 update, the regression shifted and the new fit has a tight fit and an R-squared of

-40%

-30%

-20%

-10%

0%

10%

20%

30%


% U

nd

er/O

ver

Va

lua

tio

nRio Share Price vs. Best-Fit Regression



0

10

20

30

40

50

60

70

Sh

are

Pri

ce

(£)

Rio Share Price Forecast


Anglo American: TP £20.25 (Unmodified from Trading Model Prediction)



AND LOVE THE ORE

85% (see Exhibit 269 and Exhibit 270); however, the stock is still slightly undervalued relative to its best-fit regression (see Exhibit 271). Our trading model generates a 12-month price target of £20.25, which we have taken unmodified (see Exhibit 272). Given the shift in the regression and the fundamental factors that have been weighing on the stock, the upcoming 1H reporting (at which time Mr. Cutifani will present the results of his portfolio review), and that we consider Anglo a restructuring story, we consider there is a greater likelihood of regression instability going forward than for Rio or BHP. We are monitoring the situation.



Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 270 …And Explains the Majority of the Historical Performance


0

20,000

40,000

60,000

80,000

100,000

120,000


An

glo

EV

(U

S$m

)

Anglo EV — Fitted vs. Actual

Fitted Actual

R² = 0.8536

0

20,000

40,000

60,000

80,000

100,000

120,000

0 20,000 40,000 60,000 80,000 100,000 120,000

Ang

lo A

ctu

al E

V (

US

$m

)


Anglo Scatter Plot — Fitted vs. Actual




Exhibit 271 Anglo Is Currently Slightly Undervalued Relative to Its Historical Best Fit, But We Note the Regression Has Shown Instability and a Decision to Invest in the Stock Is Contingent Upon a Belief in the Fundamental Restructuring Story as Well as Commodity Prices



-50%

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

50%

60%


% U

nd

er/O

ver

Val

uat

ion

Anglo Share Price vs. Best-Fit Regression



0

5

10

15

20

25

30

35

40

Sh

are

Pri

ce

(£)

Anglo Share Price Forecast




AND LOVE THE ORE

Vale continues to show instability in its regression relationship. This was something we had noted in our previous trading model update on May 16, 2013 as well (see European Metals & Mining: Timing Is Everything. Predictions From Our Short-Term Trading Model). At that time, we commented that the equity had previously been more than 20% undervalued at the start of the year; however, our then-updated regression found the best fit at 3% undervalued in January, and yielded a 3% discount to the May 16, 2013 price of its spot basket. Our fundamental concerns about this company cause us to question whether this regression shift may be indicative of a de-rating.

Our most recent regression analysis finds best fit at 5% overvalued in January 2013. The R-squared of the most recent update is 86% (see Exhibit 273 and Exhibit 274). At the time of our update, Vale was 8% undervalued relative to its best-fit regression (see Exhibit 275). As with Anglo American, we believe there are fundamental factors at play here. Though not a restructuring story like Anglo, this essentially pure-play iron ore company is subject to significant influence by the Brazilian government (through direct ownership of Golden Shares as well as ownership by the strategic consortium of Valepar). Furthermore, the company is, in our view, most at risk should a reduced iron ore price eventuate, given its geographic longinquity from the world’s largest consumer of iron ore, China. As a result, we consider that there is a greater likelihood of regression instability going forward than for Rio or BHP and are monitoring the situation. However, for now, we have taken the unmodified target price generated by the trading model of BRL 46.50 (see Exhibit 276).




0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

160,000

180,000

200,000

220,000

240,000


Rio

EV

(US

$m)

Vale EV — Fitted vs. Actual

Fitted Actual

Vale: TP BRL 46.50 (Unmodified from Trading Model Prediction)




Exhibit 274 …And Explains the Majority of the Historical Performance

Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 275 Vale Is Currently Slightly Undervalued Relative to Its Historical Best Fit, But as With

Anglo, We Caution That the Regression Has Been Unstable and That There Are Risks Not Captured by It (Specifically the Significant Influence of the Brazilian Government)


R² = 0.8605

50,000

70,000

90,000

110,000

130,000

150,000

170,000

190,000

210,000

230,000

250,000

50,000 70,000 90,000 110,000 130,000 150,000 170,000 190,000 210,000 230,000 250,000

Rio

Act

ual

EV

(U

S$m

)


Vale Scatter Plot — Fitted vs. Actual

-40%

-30%

-20%

-10%

0%

10%

20%

30%

40%

50%


% U

nd

er/O

ver

Va

lua

tio

n

Vale Share Price vs. Best-Fit Regression





AND LOVE THE ORE

Exhibit 276 We See Considerable Upside in the Share Price


For Glencore Xstrata, we have constructed a synthetic stock using Xstrata’s actual history and its recent relationship with Glencore. This synthetic stock shows, in our view, an acceptably high R-squared of 64% (see Exhibit 277 through Exhibit 279). We note that the difficulties of using this model and the synthetic stock for setting a target price include not only the short duration of Glencore’s trading history (having IPO'd only in May 2011) but also the distortive effects that the merger between the two companies had on their respective share prices. Furthermore, the marketing arm of Glencore (93% of 2012 revenues but 42% of operating profit that year) adds a layer of differentiation relative to the "pure" miners in our coverage. For now, we have taken the unmodified target price of £5.25 but will continue to monitor the trading model (see Exhibit 280).

0

10

20

30

40

50

60

70

80

Sh

are

Pri

ce

(£)

Vale Share Price Forecast


Glencore: TP £5.25 (Unmodified from Trading Model Prediction)




Exhibit 277 For Glencore Xstrata, the Historical Relationship Is Based on a "Synthetic" Stock Using the Xstrata Price History and Recent Relationship to Glencore

Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 278 And in Glencore Xstrata (Though for Glencore Xstrata the Historical Relationship Is

Based on a "Synthetic" Stock Using the Xstrata Price History and Recent Relationship to Glencore)


0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

160,000

180,000


Rio

EV

(US

$m)

Glencore Xstrata EV — Fitted vs. Actual

Fitted Actual

R² = 0.6421

20,000

40,000

60,000

80,000

100,000

120,000

140,000

160,000

20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 180,000

Rio

Act

ual

EV

(U

S$m

)


Glencore Xstrata Scatter Plot — Fitted vs. Actual



AND LOVE THE ORE

Exhibit 279 Glencore Xstrata Seems Fairly Valued Relative to the Historical Best Fit of This Synthetic Stock


Exhibit 280 We See Considerable Upside in the Share Price


-60%

-40%

-20%

0%

20%

40%

60%

80%


% U

nd

er/O

ver

Va

lua

tio

nGlencore Xstrata Share Price vs. Best-Fit Regression



2

3

4

5

6

7

8

9

10

May -11 Nov -11 May -12 Nov -12 May -13 Nov -13 May -14 Nov -14 May -15

Sh

are

Pri

ce

(£)

Glencore Xstrata Share Price Forecast





DCF Generates an Intrinsic Value Given the long-lived nature of mining assets, we believe a DCF is critical to understanding the intrinsic value of a share (what the share price, in our view, "ought" to be today). This is not only because we believe a DCF captures the future cash flows (and after all, what is the value of a mining company and what cash flows can be generated by its assets) but also because it forces us to consider the fundamental factors impacting the company's assets (on an asset-by-asset basis where possible, division-by-division where not) and exercise discipline — particularly important in a sector whose equity returns at least in the near term can be highly trading oriented.

Our DCF model constructed in nominal local currency terms out to 2030 over which explicit commodity price and exchange rate forecasts apply. The nominal local currency cash flows are de-escalated into real U.S. dollar cash flows and discounted at the company-specific WACC. A country risk premium reflecting the geographic origin of the cash flows is added to the underlying WACC to reflect cash flow items (i.e., expropriation) that cannot be explicitly modeled in the cash flow. All reserves are considered exploited by the model. In addition, 50% of the incremental resources (i.e., 50% of the residual resources, excluding those that have already been converted to reserves) of the company are modeled. Where residual LOM may be inferred for operations beyond the 2030 time horizon, a terminal value is calculated for the remaining years of potentially exploitable material. We use this methodology to derive all forward-looking multiples and other valuation metrics. Note that we forecast our models in reporting currency (USD), convert to listing currency (GBP or real) at an average exchange rate, and round final DCF values in 25p/cent increments.

In our call European Metals & Mining: Changing Price Targets and Upgrading Vale to Outperform, we discussed how the DCF overstates the sensitivity of mining stock prices to changes in commodity price assumptions (by a factor of 3-6x). This is in part due to the different time frames considered by market participants (ranging from a matter of days to perhaps a decade at the outside) and the DCF (which discretely models cash flows out to 2030). However, other confounding factors may be: A standard DCF model ignores the covariance between commodity prices

themselves and between commodity prices and the drivers of mining operating costs (such as FX and diesel). Such covariance between cost and price has the effect of desensitizing value to price changes.

The sensitivity analysis that is conducted using a DCF model ignores the degree of mean reversion in commodity prices. Again, this effect desensitizes value to price changes; in the extreme of instantaneous and mean reversion to a fixed long-term price, there would be no relationship between prices and mining valuations. It is worth noting that this effect applies that even if a backwardated price is used to mimic mean reversion, a backwardated curve that moves varies by a fixed quantum would show exactly the same behavior (as far as sensitivity is concerned) as a flat or contangoed price line.

Market misevaluation of risk. The final explanation is that the market places too high a discount rate on future earnings and too low a discount rate on current earnings. This has the same effect of increasing the apparent degree of commodity price backwardation in the miners and would reduce the sensitivity of the equity to changes in the commodity price basket.

DCF Generates an Intrinsic Value for the Stock Today, But Is Silent on When the Stock May Achieve That Value



AND LOVE THE ORE

Exhibit 281 BHP DCF Value

Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 282 Rio Tinto DCF Value


US$m DCF Terminal Value Total Per Share (US$) Per Share (£) 2013 EBITDA EV/EBITDA

Petroleum 90,058 - 90,058 16.92 11.10 11,592 7.8

Aluminum 5,089 942 6,032 1.13 .74 165 36.5

Base Metals 39,562 8,147 47,709 8.97 5.88 4,689 10.2

Diamonds & Specialty Products (3,822) - (3,822) (.72) (.47) 19 (202.1)

Stainless Steel Materials (656) - (656) (.12) (.08) (37) 17.8

Iron Ore 82,461 14,654 97,115 18.25 11.97 14,271 6.8

Manganese 3,252 56 3,308 .62 .41 514 6.4

Metallurgical Coal 11,163 3,639 14,803 2.78 1.83 1,395 10.6

Energy Coal 5,971 1,174 7,145 1.34 .88 777 9.2

BHPB Enterprise Value 233,079 28,613 261,691 49.18 32.27 33,384 7.8

Less Minorities (1,858) (.35) (.23)

Less Central Costs (9,328) (1.75) (1.15)

Less (Net Debt)/Net Cash (32,169) (6.05) (3.97)

Equity Value 218,337 41.03 26.92

Shares Outstanding 5,321

Share Price (USD) 41.03

Exchange Rate 1.52

Implied Share Price (£) 26.92


Iron Ore 97,333 18,889 116,222 62.94 41.29 18,801 6.2

Aluminum 14,801 6,522 21,323 11.55 7.58 1,326 16.1

Copper 37,448 8,266 45,715 24.76 16.24 2,016 22.7

Energy 8,709 2,027 10,736 5.81 3.81 1,111 9.7

Diamonds & Minerals 8,213 - 8,213 4.45 2.92 1,290 6.4

Other (3,070) - (3,070) (1.66) (1.09) (360) 8.5

Rio Enterprise Value 163,434 35,705 199,139 107.85 70.75 24,183 8.2

Less Minorities (21,423) (11.60) (7.61)

Less Central Costs - - -


Equity Value 155,726 84.34 55.33



Exchange Rate 1.52

Implied Share Price (GBP) 55.33




Exhibit 283 Anglo American DCF Value

Source: Bloomberg L.P., corporate reports and Bernstein estimates and analysis. Exhibit 284 Vale DCF Value



Platinum 4,565 1,149 5,714 4.47 2.93 920 6.2

Diamonds 5,375 - 5,375 4.21 2.76 1,498 3.6

Copper 20,118 3,002 23,120 18.09 11.87 2,234 10.3

Nickel 1,204 227 1,432 1.12 .73 130 11.0

Iron Ore — South Africa 16,931 1,760 18,691 14.63 9.59 3,507 5.3

Iron Ore — Brazil 4,599 1,610 6,209 4.86 3.19 228 27.2

Manganese 2,550 - 2,550 2.00 1.31 445 5.7

Metallurgical Coal 2,378 373 2,751 2.15 1.41 263 10.4

Thermal Coal 4,950 229 5,179 4.05 2.66 721 7.2

Other — Non-Mining Assets 3,030 - 3,030 2.37 1.56 486 6.2

Anglo Enterprise Value 65,700 8,351 74,051 57.94 38.01 8,815 8.4

Less Minorities (20,174) (15.79) (10.36)



Equity Value 45,653 35.72 23.44



Exchange Rate 1.52


US$m DCF Terminal Value Total Per Share (US$) Per Share (BRL) 2012 EBITDA EV/EBITDA

Iron Ore 88,097 31,317 119,414 23.32 47.00 17,647 6.8

Manganese & Ferroalloys 225 - 225 .04 .09 190 1.2

Coal (1,316) 1,295 (21) (.00) (.01) (449) .0

Base Metals 21,437 2,397 23,833 4.66 9.38 603 39.5

Fertilizers 9,240 - 9,240 1.80 3.64 632 14.6

Other (4,884) - (4,884) (.95) (1.92) (490) 10.0

Vale Enterprise Value 112,800 35,009 147,808 28.87 58.18 18,133 8.2

Less Minorities (3,398) (.66) (1.34)



Equity Value 118,322 23.11 46.57



Exchange Rate 2.02

Implied Share Price (BRL) 46.57



AND LOVE THE ORE

Exhibit 285 Glencore Xstrata DCF Value (With and Without Synergies)



Zinc 13,146 1,760 14,907 1.12 .70 2,000 7.5

Copper 51,956 6,558 58,513 4.39 2.74 4,595 12.7

Nickel 7,583 1,395 8,978 .67 .42 663 13.5

Alumina & Aluminum 384 - 384 .03 .02 13 30.5

Coal 17,586 3,098 20,684 1.55 .97 3,106 6.7

Oil & Oil Products 3,353 284 3,636 .27 .17 419 8.7

Agricultural Products 755 - 755 .06 .04 142 5.3

Alloys (75) - (75) (.01) (.00) 144 (.5)

Marketing — Metals & Mining 8,382 1,288 9,670 .73 .45 1,289 7.5

Marketing — Energy Products 5,479 971 6,450 .48 .30 717 9.0

Marketing — Agricultural Products 5,481 892 6,372 .48 .30 220 29.0

Synergies 4,241 760 5,001 .38 .23 750 6.7

With Synergies

Glencore Enterprise Value 118,270 17,006 135,276 10.15 6.33 14,058




Plus Corporate & Other

Equity Value 93,739 7.03 58,441.02


Per Share NPV (USD) 7.03

Exchange Rate 1.60

Per Share NPV (GBP) 4.39

Without Synergies






Equity Value 88,739 6.66 58,216.90



Exchange Rate 1.52





Risks to Our Sector and Stocks The four most significant risks facing the major mining houses in Europe are lack of capital discipline, operating cost inflation, a sustained downturn in the Chinese economy and resource nationalism. Capital discipline: We believe that mined commodity prices will stay high and

will continue to trend higher until such point that the massive amounts of labor currently employed in the Chinese mining industry are displaced by capital. This can happen either through a reform of the domestic mining industry in China or through the displacement of that industry by supply increases in other geographies. We do not believe that the natural resource endowment of China will allow for a rapid (if any) domestic reform, as this requires the existence of massive high-grade, long-life deposits (such as copper in Chile or iron ore in Brazil and Australia). Consequently, the duration of the current pricing environment comes down to the extent to which the Western capital deployed by the major mining houses to increase low-cost commodity production will displace the requirement for Chinese domestic production.

Operating cost inflation: U.S. dollar denominated unit costs in all the major mining houses have seen double-digit growth rates over the last 10 years. Part of this can be explained by movements in exchange rates and part by the prevailing inflationary environment in producer geographies. However, there has still been a very significant increase in underlying real costs. Should this persist or accelerate, then it has the capacity to erode value.

Chinese economic risks: China is important in commodities as both the major source of demand growth and as the location of the marginal units of supply. We believe that the current slowdown in the Chinese economy has been largely self-induced in an effort to contain food inflation and, as a consequence, a political stasis ahead of the handover of power at the end of the year. However, a more long-lived slowdown has the capacity to move the commodity markets into oversupply on a sustained basis — particularly if new supply is not curtailed.

Resource nationalism: Finally, we note with concern the trend toward global fragmentation and the ever greater desire to extract value from the mining sector. We believe that this is ultimately a self-defeating strategy by host governments but it is one with an impressively long pedigree. Persistent macroeconomic headwinds will make this an ever more attractive option.

Rio Tinto PLC: Company-specific risks include any sustained downturn in iron ore prices, as the company is the second most exposed of our coverage group to iron ore (after Vale). Any relaxation of capital discipline particularly around the Simandou project in Guinea would also be, in our view, a negative catalyst. Further execution delays in the commissioning of the Oyu Tolgoi copper project in Mongolia or significant revenue grabs from the Mongolian government could also be a risk.

BHP Billiton PLC: Company-specific risks include continued weakness in the price of natural gas in the U.S. and iron ore. Repeats of the weather-induced volume losses in BHP’s metallurgical coal operations as well as continued labor-related disruptions in these assets could also prove an impediment to our target price.

Vale SA: Company-specific risks include any sustained downturn in iron ore price, as the company derives nearly the entirety of its value from exposure to iron ore. The continuation of disruptions to the output of nickel and attendant cost pressures are also a risk. The commissioning of Goro (VNC) is an issue that needs to be addressed, as is the performance at Onca Puma.

Anglo American PLC: For Anglo American in particular, inability to improve the efficiency of its platinum operations and continued margin pressure arising from South African labor inflation poses downside risk, as does the potential for increased union militancy in South Africa (and again in platinum, in particular). A continued deterioration in labor unrest along with the attendant physical hazards, delays and expenses could weigh on results. Further delays at the Minas Rio iron

Risks



AND LOVE THE ORE

ore project in Brazil would also be a significant negative catalyst. Failure to properly integrate De Beers into the Anglo American portfolio could again risk value loss.

Glencore Xstrata PLC: The greatest unknown, in our minds, for Glencore relates to its marketing activities. We believe there is insufficient transparency to assess both embedded risk in the trading book and persistence of edge. We also note that Glencore requires high levels of working capital and remains vulnerable to large swings in cash-flow generation as a result. We note as a result of its operations in frontier jurisdictions, as well as the unknown nature of embedded risk and persistence of edge in the marketing book, headline risk remains a significant concern for Glencore. Lastly, we do not consider Glencore yet "institutional quality" in terms of its transparency or governance.

Post merger, we see challenges facing the combined Glencore-Xstrata entity, specifically integration of Xstrata's significantly larger operating business into Glencore's management structures while avoiding disruption to a number of critical projects, particularly given the anticipated dismissal of many Xstrata personnel. Furthermore, the choice for a new chairman will be critical to ensuring that the corporate governance and minority shareholder interests are well protected as Glencore enters a new stage in its life as one of the largest publicly listed mining companies in the world.




Company Impact

Iron ore is the single most important commodity in the mining space, representing 58% of EBITDA generation (see Exhibit 286). Furthermore, it has been one of the strongest performing commodities throughout the mining "super-cycle" (see Exhibit 287), having risen by 1,340% from its 2001 average price to its March 2011 peak.44 Consequently, having a view on this commodity is essential for understanding the bulk of the value contained within and generated by our coverage group (see Exhibit 288).45

Although it represents the lion's share of our coverage EBITDA generation, iron ore's importance to each individual miner varies substantially across companies. Vale and Rio Tinto have the largest exposure, with iron ore contributing 97% and 75% to their 2012 EBITDA, respectively. As the most diversified miners, Anglo American and BHP Billiton saw 37% and 55% of EBITDA, respectively, coming from the cold metal. Finally, mining operations of Glencore Xstrata have no direct production exposure to iron ore bar the 1Q:13 acquisition of a minority stake in Ferrous Resources.

This chapter is divided into three sections: highly concentrated iron ore plays (Rio and Vale), diversified miners with moderate exposure (Anglo American and BHP Billiton) and companies with no direct production exposure (Glencore Xstrata).

Exhibit 286 Iron Ore Represents, by Far, the Most Significant Contribution to the Cash Generation of the Miners…

Exhibit 287 …And Has Been, Along With Copper, the Strongest Performer of the Major Commodities

Source: Corporate reports and Bernstein analysis. Source: Bloomberg L.P. and Bernstein analysis.

44 The IMF 62% Fe monthly spot (CFR Tianjin port), 2001 average = US$13/t, March 2011 = US$187/t. 45 Glencore Xstrata is the only company in our coverage without any direct production exposure to iron ore.

58%

18%

8%

15%2012 EBITDA Breakdown

Iron Ore Base Metals Coal Other

0

100

200

300

400

500

600

700

800

900

1000

Jun

e 2

000

= 10

0

Indexed Commodity Price Performance

Iron Ore Copper HCC

Iron Ore Currently Accounts for ~60% of the EBITDA of the Five Major Mining Houses; It Has Been the Outperforming Commodity in the Last 10 Years



AND LOVE THE ORE

Exhibit 288 Vale Represents Essentially an Iron Ore Pure Play in Both the Near and Longer Term, While We See Rio Diversifying in the Medium Term; BHP and Anglo American Have Exposure to a Well-Diversified Basket of Commodities; Glencore Xstrata Is the Only Company in Our Coverage Without Any Direct Production Exposure to Iron Ore*

* Excluding the 1Q:13 acquisition of a minority stake in Ferrous Resources.

Source: Corporate reports and Bernstein analysis.

Vale and Rio Tinto — Highly Concentrated Iron Ore Plays In our coverage group, Vale and Rio Tinto are the most heavily exposed to iron ore (97% and 93% of 2012 EBIT, respectively).

We consider Rio Tinto the most attractive stock in our coverage. Not only does the company have significant iron ore exposure (93% of 2012 EBIT), it does not suffer from the same risks — quantifiable and unquantifiable — that we see impacting Vale (which generated 97% of its 2012 EBIT from iron ore). Exhibit 289 and Exhibit 290 illustrate the 2012 breakdown of Rio Tinto's and Vale's revenues and operating profits by commodity. Exhibit 291 through Exhibit 294 show a historical evolution of revenues and EBITDA by commodity since 2001.

0%

20%

40%

60%

80%

100%

120%

Anglo BHP Rio Vale Glencore Xstrata

Contribution of Iron Ore to EBITDA (2012)

Iron Ore Other

Both Rio and Vale Derive the Bulk of Their Earnings (93% and 97% of 2012 EBIT, Respectively) from Iron Ore; We Prefer Rio, Which Also Benefits from World-Class Copper Exposure




Exhibit 289 Vale Is the Most Heavily Exposed to Iron Ore of Our Coverage Group (a Trend We Do Not See Changing)…

Exhibit 290 …Followed by Rio Tinto, Which Is Augmenting Its Portfolio With Some of the World's Best Copper Projects (the Metal on Which We Are the Most Bullish)

Source: Corporate reports and Bernstein analysis. Source: Corporate reports and Bernstein analysis.

Exhibit 291 Historical Evolution of Rio Tinto's Revenue by Business Unit


70%

-4%

1%3%

8%

4%

15%

97%

2%1%

100%

EBIT

3%

0% 2%

-2%

Group Revenue

Others

Manganese & Ferroalloys

Coal

Logistics

Fertilizers

Base Metals

Iron Ore

Vale

7%

-7%

10%

9%4%

12%

7%

100%

EBIT

93%

Group Revenue

3%

0%

18%

44%Other

Diamonds & Minerals

Energy

Copper

Aluminium

Iron Ore

Rio Tinto

-

10,000

20,000

30,000

40,000

50,000

60,000

70,000

2001 A 2002 A 2003 A 2004 A 2005 A 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A

US

$m

Revenue by Business Unit

Iron Ore Aluminum Copper Energy Diamonds & Minerals Other



AND LOVE THE ORE

Exhibit 292 Historical Evolution of Rio Tinto's EBITDA by Business Unit


Exhibit 293 Historical Evolution of Vale's Revenue by Business Unit


-

5,000

10,000

15,000

20,000

25,000

30,000

35,000

2001 A 2002 A 2003 A 2004 A 2005 A 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A

US

$mEBITDA by Business Unit


-

10,000

20,000

30,000

40,000

50,000

60,000

70,000

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

US

$m


Iron Ore Manganese & Ferroalloys Coal Base Metals Fertilizers Logistics




Exhibit 294 Historical Evolution of Vale's EBITDA by Business Unit


Vale is the world's largest iron ore producer and, in Carajas, has one of the

world's most attractive iron ore assets. However, it is the most operationally geared miner to the iron ore price (partially as a consequence of geographical distance to China, partially as a consequence of the operating cost position of some of the Southern mines and the pellet operations, and partially as a consequence of its significant capital spend). Consequently, any change in the long-term price of iron ore has a huge impact on the DCF value of Vale. Moreover, we also believe that there is more to the story of iron ore prices than simple monotonic declines to some low flat price. Consequently, there is still an asymmetric risk to the upside for iron ore prices in the short term. So from a pure value consideration, we are more favorably inclined to Vale now than previously. Moreover, it has a diversification strategy, which is beginning to move the company away from being a single-commodity pure play and most importantly increasing the company's exposure to copper with Salobo and Lubambe.

That said, we still favor Rio Tinto for those wishing to gain iron ore exposure. Rio Tinto owns some of the highest-quality iron ore assets and (vitally) infrastructure globally. Rio has offered a slightly better "bang for your iron ore buck" on correlation and absolute return bases — at less than half the capex per incremental ton of production and without the risk of significant involvement by the Brazilian government. Rio also has exposure to some of the world's best operational copper assets — not to mention three of the world's best undeveloped copper deposits. Tier 1 assets in iron ore include Dampier and Cape Lambert (Australian Pilbara), while copper includes Escondida, Grasberg and Bingham Canyon. The company also has some of the most promising and largest copper prospects in Oyu Tolgoi, Resolution and La Granja.

Rio has commodity diversification into copper and mineral sand, rather than nickel, and has demonstrated an ability to deliver iron ore growth at considerably lower capital intensity than Vale. That, coupled with the prospects of genuine capital discipline and cost cutting under new CEO Sam Walsh (who made his bones in low-cost brownfield Pilbara production) and a more reasoned approach to volume growth, means that we continue to see Rio Tinto as our top pick. Exhibit 295 through Exhibit 298 show our revenue and EBITDA forecasts for Rio Tinto and Vale.

-

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

US

$mEBITDA by Business Unit

Iron Ore Manganese & Ferroalloys Coal Base Metals Fertilizers Logistics



AND LOVE THE ORE

Exhibit 295 Our Forecasts for Rio Tinto's Revenues… Exhibit 296 …And EBITDA Through 2018 See Considerable Growth


Exhibit 297 Similarly, Our Forecasts for Vale's Revenues…

Exhibit 298 …And EBITDA Also Sees Considerable Growth


-10,00020,00030,00040,00050,00060,00070,00080,00090,000

100,000

Re

ven

ue

(US

$m)

Rio Tinto Revenue


(10,000)

-

10,000

20,000

30,000

40,000

50,000

EBIT

DA

(U

S$m

)

Rio Tinto EBITDA


-

10,000

20,000

30,000

40,000

50,000

60,000

70,000

80,000

2011

2012

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

Re

ven

ue

(US

$m)

Vale Revenue

Iron Ore Manganese & Ferroalloys Coal Base Metals Fertilizers Logistics Others

-

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,00020

11

2012

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

EBIT

DA

(U

S$m

)

Vale EBITDA

Iron Ore Manganese & Ferroalloys Coal Base Metals Fertilizers Logistics Others




BHP Billiton and Anglo American — The Diversified Miners BHP Billiton not only has the longest corporate history of our coverage group (its component parts trace their origins back to 1851 and 1883, respectively), we view it as the highest quality stock in our coverage. In 2012, the company generated the highest revenue of the "Big Three" miners in our coverage (US$72 billion versus US$56 billion for Rio and US$49 billion for Vale) and had the highest EBITDA margin (42% versus 39% for Vale and 36% for Rio).

BHP Billiton has a well-diversified portfolio from a commodity exposure perspective. Lagging only Anglo American, it is the second most diversified company in our coverage (see Exhibit 299 through Exhibit 302). Both Anglo and BHP have exposure to a commodity that behaves differently from either the ferrous or the base metals. For Anglo, it is platinum, and for BHP Billion, it is petroleum. However, where platinum has been a drag on Anglo's portfolio, BHP has received a boost from "black gold" related in its petroleum division (69% EBITDA margins in calendar 2012 versus 11% for Anglo's platinum division) — Exhibit 303 and Exhibit 304. We attribute BHP's high margins relative to peers to the Tier 1 nature of the company’s assets and the operational excellence with which they are run. Consequently, it is BHP Billiton that is the most diversified miner from an earnings perspective.

The strong commodity risk diversification of BHP Billiton is complemented with geographic diversification that, while skewed to Australia, is nonetheless low risk. A significant portion of BHP's assets is located either in Australia (iron ore, coal and petroleum), North America (petroleum) or Chile (copper). There is limited emerging market asset exposure and, consequently, limited political risk in the portfolio (see Exhibit 300). That said, we have noted with concern the deterioration in the attractiveness of Australia as a location for new mining investment over the last few years. We believe that this provides a reminder that no country is ever entirely "safe." Nevertheless, we continue to see significant revenue and EBITDA growth for "The Big Australian" (see Exhibit 305 and Exhibit 306).

Exhibit 299 The Diversity in Exposure Makes BHP One of the Lowest-Risk Miners from a Commodity Perspective…

Exhibit 300 …Complemented by Low Political Risk and a Diverse Range of Operating Geographies


7%

9%

19%

26%

19%

20%

29%

52%

0%

3%

0%

4%

0%

8%

3%

100%

EBIT

-2%1%

Group Revenue

1%

Diamonds

Manganese

Stainless Steel Materials

Aluminum

Energy Coal

Metallurgical Coal

Petroleum

Base Metals

Iron Ore

BHP Billiton

55%

0%

15%

13%

8%1%

8%

Net Assets by Location

Australia UK North America South America

Southern Africa Rest of World Unallocated

BHP Billiton Is the Highest Quality Mining Stock in Our Coverage, With Genuine Tier 1 Assets (e.g., Escondida, Mt Newman, Olympic Dam and Peak Downs) Diversified Across a Range of Commodities and With Limited Geographic Risk



AND LOVE THE ORE

Exhibit 301 Historical Evolution of BHP's Revenue by Business Unit


Exhibit 302 Historical Evolution of BHP's EBITDA by Business Unit


-

10,000

20,000

30,000

40,000

50,000

60,000

70,000

80,000

2001 A 2002 A 2003 A 2004 A 2005 A 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A

Re

ven

ue

(US

$m)


Petroleum Aluminum Base Metals Diamonds & Specialty ProductsStainless Steel Materials Iron Ore Manganese Metallurgical CoalEnergy Coal Other Group & Inter Segment

-

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

45,000

2001 A 2002 A 2003 A 2004 A 2005 A 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A

EB

ITD

A (U

S$m

)

EBITDA by Business Unit

Petroleum Aluminum Base Metals

Diamonds & Specialty Products Stainless Steel Materials Iron Ore

Manganese Metallurgical Coal Energy Coal




Exhibit 303 BHP's Petroleum Exposure Is Its Great Diversifier (Generating 69% EBITDA Margins in 2012)…

Exhibit 304 …While Anglo American's Great Diversifier, Platinum, Came in at Margins of Just 11%


Exhibit 305 We Also See Revenue and EBITDA Growth…

Exhibit 306 …For BHP Billiton


Anglo American has the most diversified portfolio of our coverage group from a commodity perspective and is significantly smaller than the "Big Three" of BHP, Rio and Vale. There is a distinct reduction in scale when stepping down from this triumvirate of the "Big Three" miners to Glencore Xstrata and Anglo (see Exhibit 307).46 While it is tempting to talk of the five largest global miners as peers, we believe that the gulf between them is now so large as to make that unconvincing.

The reason for the emergence of the "Big Three" and for the relative ranking in EBITDA margin performance is attributable to iron ore. Those miners that have

46 We believe that Glencore Xstrata would need to merge with Anglo American to create a competitor to the "Big Three."

-

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

-

20

40

60

80

100

120

140D

ec-0

1D

ec-0

2D

ec-0

3D

ec-0

4D

ec-0

5D

ec-0

6D

ec-0

7D

ec-0

8D

ec-0

9D

ec-1

0D

ec-1

1D

ec-1

2

US

$m

Pe

tro

leum

Pro

duct

ion

(mm

boe)

Petroleum

Petroleum Production (LHS) Petroleum Revenue (RHS)

Petroleum Capex (RHS)

-

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

-200

400600

8001,000

1,2001,4001,600

1,800

Jun-

01Ju

n-02

Jun-

03Ju

n-04

Jun-

05Ju

n-06

Jun-

07Ju

n-08

Jun-

09Ju

n-10

Jun-

11Ju

n-12

US

$m

Pt P

rod

uct

ion

(ko

z)

Platinum

Platinum Production (LHS) Platinum Revenue (RHS)

Platinum Capex (RHS)

-

20,000

40,000

60,000

80,000

100,000

120,000

140,000

2011 A

2012 A

2013 E

2014 E

2015 E

2016 E

2017 E

2018 E

Re

ven

ue

(US

$m)

BHP Billiton Revenue

Petroleum AluminumBase Metals Diamonds & Specialty ProductsStainless Steel Materials Iron Ore

Manganese Metallurgical CoalEnergy Coal Other Group & Inter Segment

(10,000)

-

10,000

20,000

30,000

40,000

50,000

60,000

70,000

2011 A

2012 A

2013 E

2014 E

2015 E

2016 E

2017 E

2018 E

EBIT

DA

(U

S$m

)

BHP Billiton EBITDA

Petroleum Aluminum

Base Metals Diamonds & Specialty Products

Stainless Steel Materials Iron Ore

Manganese Metallurgical Coal

Energy Coal

Anglo American Has the Most Diversified Portfolio in Our Coverage — It Remains a Turnaround Story in Our Minds



AND LOVE THE ORE

exposure to this commodity enjoy the highest margins and have seen the fastest revenue growth. To the extent that the iron ore margins are diluted by other less well-performing commodities, the lower the overall EBITDA margin will be (see Exhibit 308). In general, Anglo benefits from having a well-diversified portfolio of different commodity exposures (see Exhibit 309, Exhibit 311 and Exhibit 312). In this regard, it is probably the most diversified of the companies in our coverage and has the lowest pure commodity risk. This lower commodity risk is, however, partially offset by greater country risk. Anglo remains heavily exposed to South Africa through Kumba, Anglo Platinum and its thermal coal operations. We also see the company becoming increasingly reliant on South America for future earnings growth. Geographically, Anglo American lacks a really significant presence in Australia, with antipodean exposure confined to metallurgical coal (a key differentiator from BHP Billiton, the second most diversified of our coverage companies from a commodity perspective) — Exhibit 310.

The performance at Anglo Platinum is of particular concern to us, as the overall level of flat production belies a significant shift in the source of the material that stands behind this. Following the requirements of South African Black Economic Empowerment (BEE), we see significant material now coming via concentrate purchases from JV partners and associates (see Exhibit 314). Only ~75% of Anglo Platinum's refined output comes from its own mined material (see Exhibit 313). This change has a significant impact on margins, as we believe that Anglo Platinum purchases material for about 90% of the price of the contained metal. Given that refining costs must still be borne, we see purchases of material as being dilutive of the margins that can be achieved from mining.

Anglo remains in our minds a turnaround story — one that will challenge new CEO Mark Cutifani, but one which we believe is solvable for a leader with his 36 years of operational expertise. We are looking forward to the results of his strategic review.47 As an operationally experienced miner across a host of commodities and geographies, including deep precious metal mining in Africa, we believe he is well prepared to address the issues currently facing Anglo Platinum and the group as a whole. Going forward, we will be particularly interested to hear more from the company about the progress of its Platinum business restructuring and the development of Minas Rio, and we expect to hear more details in his portfolio review, to be presented at the upcoming half-yearly reporting. Exhibit 315 and Exhibit 316 show our revenue and EBITDA forecasts.

47 Anglo's upcoming reporting schedule is as follows: 2Q production reports for Anglo and listed subsidiaries Kumba and Anglo Platinum are scheduled for Thursday, July 18, 2013. Anglo Platinum's 1H earnings report is scheduled for Monday, July 22, 2013; Kumba's for Tuesday, July 23, 2013; and Anglo American's for Friday, July 26, 2013. All reporting times 7am BST.




Exhibit 307 Anglo and Glencore Xstrata Represent the Tier 2 of Mining Companies Behind the "Big Three" of Vale, Rio and BHP

Exhibit 308 The Differentiating Feature of the "Big Three," Both in Terms of Size and Margin, Is the Presence of Significant Iron Ore Exposure

Note: Only the revenue and EBITDA for Glencore's Mining and Energy Industrial activities are included.


Note: Only the revenue and EBITDA for Glencore's Mining and Energy Industrial activities are included.


Exhibit 309 Anglo Is the Most Diversified and the Lowest-Risk Miner from a Commodity Exposure Perspective

Exhibit 310 The Relative Lack of "Safe" Australian Exposure Has Always Been a Risk Compared to the Other Miners


0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

80,000

Re

ven

ue

(US

$m) D

ece

mbe

r Ye

ar E

nd

Sector Revenue and EBITDA

Revenue EBITDA

0%

5%

10%

15%

20%

25%

30%

35%

40%

Vale Rio Tinto BHP Billiton

Anglo Glencore Xstrata

Sector EBITDA Margin

12%

7%

12%

8%

11%

13%

0%

100%

EBIT

-3%-2%

27%

49%

Group Revenue

11%

17%

1%

16%

21%

Other

Platinum

Nickel

Metallurgical Coal

Diamonds

Thermal Coal

Copper

Iron Ore and Manganese

Anglo American

36%

1%46%

2%

10%

5%Net Assets by Location

South Africa Other Africa South America

North America Australasia Other



AND LOVE THE ORE

Exhibit 311 Historical Evolution of Anglo American's Revenue by Business Unit


Exhibit 312 Historical Evolution of Anglo American's EBITDA by Business Unit


-

5,000

10,000

15,000

20,000

25,000

30,000

35,000

40,000

45,000

2001 A 2002 A 2003 A 2004 A 2005 A 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A

US

$mRevenue by Business Unit

Platinum Diamonds Copper Nickel Iron Ore and Manganese Metallurgical Coal Thermal Coal Other Mining & Industrial Other

-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

2001 A 2002 A 2003 A 2004 A 2005 A 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A

US

$m

EBITDA by Business Unit

Platinum Diamonds Copper Nickel Iron Ore and Manganese Metallurgical Coal Thermal Coal Other Mining & Industrial Other




Exhibit 313 An Increasing Proportion of Anglo Platinum's Refined Production Now Comes from Non-Mined Sources…

Exhibit 314 …The Majority Is Purchased from Related Joint Venture Partners and Associates


Exhibit 315 Our Forecast for Anglo American Sees Revenue Growth…

Exhibit 316 …As Well as EBITDA Growth as Part of the Turnaround Story


0200400600800

1,0001,2001,4001,6001,800

Jun-

2001

Jun-

2002

Jun-

2003

Jun-

2004

Jun-

2005

Jun-

2006

Jun-

2007

Jun-

2008

Jun-

2009

Jun-

2010

Jun-

2011

Jun-

2012

Jun-

2013

ko

z P

t

Refined Platinum Production by Source

Series1 Series2

74%

11%

11%4%

Refined Pt by Source, 2012

Refined Pt from Mined Production Conc from JVs

Conc from Assocs Conc from Third Party

-

10,000

20,000

30,000

40,000

50,000

60,000

2010 A

2011 A

2012 A

2013 E

2014 E

2015 E

2016 E

2017 E

2018 E

Re

ven

ue

(US

$m)

Anglo American Revenue

Platinum Diamonds Copper

Nickel Iron Ore and Manganese Metallurgical Coal

Thermal Coal Other Mining & Industrial Other

-

5,000

10,000

15,000

20,000

25,000

2010 A

2011 A

2012 A

2013 E

2014 E

2015 E

2016 E

2017 E

2018 E

EBIT

DA

(U

S$m

)

Anglo American EBITDA

Platinum Diamonds Copper

Nickel Iron Ore and Manganese Metallurgical Coal

Thermal Coal Other Mining & Industrial Other



AND LOVE THE ORE

Glencore Xstrata Has No Direct Iron Ore Production Exposure Glencore Xstrata offers exposure for copper and coal bulls without a taste for iron (given that the company's only direct iron ore exposure is its 3Q:13 acquisition of a stake in Ferrous Resources). Glencore Xstrata also offers exposure to the sales and trading house that CEO Ivan Glasenberg built. As such, it has a profile that is distinct relative to the pure miners in our coverage (see Exhibit 317 and Exhibit 318). The impact of the high-turnover, low-margin trading business (2% EBITDA margin in 2012 versus 25% for Glencore Xstrata combined) is even clearer when contextualized against the average 33% for our coverage group (see Exhibit 319 and Exhibit 320).

On the industrial-assets side, Glencore Xstrata is predominantly a copper and coal play, with zinc a tertiary significant exposure (see Exhibit 321 through Exhibit 323). The copper and coal assets from the erstwhile Xstrata portfolio contain some particularly compelling operations. In Latin American copper, Collahuasi, Antamina and Alumbrera are particularly attractive operating assets in our view, while in coal, we consider Okay Creek, Rolleston, Bulfa, Mangoola, Ulan (all Australia) and Cerrejon (Columbia) to be gems.

It is worth noting that, from a strategic perspective, we prefer pure mining sector plays to diversified natural resources companies (hence our positive view on the strategic repositioning that Anglo American has undertaken). To the extent that mining companies have energy — specifically, oil and gas — production in their portfolios and can run those at similar or better operating profit margins to the oil and gas majors, we are not unduly troubled by the inherent diversification from mining (despite the different skill sets and technical expertise required by the two disciplines). Agriculturals are, however, not an obvious fit to us, in terms of operational skill sets, time horizons for management, margins, etc. We understand Glencore’s historical reasons for developing this portion of its asset portfolio, growing as it did from a trading company. However, we remain concerned that the diversity of the group splits management's focus, diverting energy and attention away from the most attractive mining elements of the portfolio. We see this as a negative for the company.

Copper- and Coal-Heavy Xstrata Is the Miner for Sector Bulls Without a Taste for Iron — But One Has to Have An Appetite for Headline Risk




Exhibit 317 Glencore Xstrata's High-Revenue, Low-Margin Trading Business Renders Its Group Profile Very Different from the Pure Miners in Our Coverage…

Exhibit 318 …Though Excluding the Marketing Division Reveals a Profile That Is Similar to the Rest of the Miners


Note: Only the revenue and EBITDA for Glencore Xstrata's Industrial Activities are included.


Exhibit 319 The Impact of the Marketing Division Can Be Seen in Total Company Margins…

Exhibit 320 …While the Margins for Only Industrial Activities More Closely Resemble the Pure Miners in Our Coverage


0

50,000

100,000

150,000

200,000

250,000

Re

ven

ue

(US

$m) D

ece

mbe

r Ye

ar E

nd

Sector Revenue and EBITDA, 2012

Revenue EBITDA

010,00020,00030,00040,00050,00060,00070,00080,000

Rev

enu

e (U

S$m

) Dec

emb

er Y

ear E

nd

Sector Revenue and EBITDA, 2012

Revenue EBITDA

0%

5%

10%

15%

20%

25%

30%

35%

40%

Sector EBITDA Margin, 2012

0%

5%

10%

15%

20%

25%

30%

35%

40%


Anglo Glencore Xstrata

Industrial Activities

Sector EBITDA Margin, 2012



AND LOVE THE ORE

Exhibit 321 Historical Pro Forma Evolution of Glencore Xstrata's Industrial Activities EBITDA


Exhibit 322 Glencore Xstrata's Industrial Activities Are Predominantly a Copper, Coal and Zinc Play…

Exhibit 323 …As Can Also Be Seen When Looking at the Company in Toto (Inclusive of Marketing)

Note: Only Industrial Activities shown, marketing and associates have been omitted.



We consider the recently completed merger between Glencore and Xstrata to

be the logical conclusion of a journey nearly two decades in the making, one that was kicked off by the 1994 management buyout at Glencore and which catapulted Mr. Ivan Glasenberg into a senior role at the company he now runs. Mr. Glasenberg has proved to be a quick learner in his transition from a prominent-yet-private

(500)

1,500

3,500

5,500

7,500

9,500

11,500

13,500

15,500

2008 A 2009 A 2010 A 2011 A

Glencore-Xstrata Pro Forma EBITDA — Industrial Activities (US$mln)

Zinc Copper Nickel Alumina & Aluminum Coal Oil & Oil Products Agricultural Products Alloys Associates Total

Zinc, 18%

Copper, 41%

Nickel, 6%

Alumina & Aluminum, 0%

Coal, 28%

Oil & Oil Products, 4%

Agricultural Products, 1%

Alloys, 1%

Glencore Xstrata Pro Forma EBITDA Industrial Activities (2012)

Zinc, 15%

Copper, 36%

Nickel, 8%Alumina & Aluminum, 0%

Coal, 27%

Oil & Oil Products, 0%

Agricultural Products, 0%

Alloys, 2%

Marketing, 17%

Associates, 1%

Other, -7%

Glencore Xstrata Pro Forma EBITDA (2012)




master trader to the CEO of a publicly listed company — as evidenced to investors who watched him navigate 2H:12 demands from shareholders for an improved share ratio and alterations to management compensation packages. His decision to prolong this merger process in order to "negotiate" with a regulator that does not oversee direct operations but represents the interests of his largest consumer showed an awareness of the importance of customer relationship management. The concessions that he granted addressed the concerns of the Chinese and left Mr. Glasenberg several options to unload greenfield projects that he was not particularly interested in pursuing — or the disposal of an admittedly costly crown jewel at a potentially lucrative price.

We hope that Mr. Glasenberg will show an equally steep learning curve when it comes to transparency of reporting and governance. We consider that Glencore has some way to go before it will be "institutional quality" on these metrics. Furthermore, potential investors should note that Glencore Xstrata's country risk premium is the second highest in our coverage group (2%, behind only Anglo American's 2.6% versus an average of 1.2% for the "Big Three"). This is a function of the company's operations in frontier jurisdictions like the DRC. However prudently one conducts business in these jurisdictions, they carry not only an elevated risk that host governments may make increased grabs for a share of the revenues, but also headline risk — as does Glencore's trading division.48

Nevertheless, as an owner operator (he holds 8% of the combined entity's paper), Mr. Glasenberg's incentives are squarely aligned with investors — indeed, media coverage of his "rich" 2012 compensation package seemed to fail to grasp what a good thing it is indeed that the bulk of his compensation comes from company dividends. We have long been calling for capex restraint to support commodity prices, and have highlighted how increased dividends can not only offer a boost to the sector in terms of supporting commodity prices, but also can lead to equity rerating. It was appropriate for the sector to pursue an aggressive growth strategy in the last half of the last century; this century brings a new environment to which mining companies must respond if they are to be responsible custodians of the value created by previous generations of management.49 Mr. Glasenberg seems to be one of the few industry leaders who truly understand this.

It remains to be seen how he will staff the combined entity, following the departure of much of the senior management talent from Xstrata. But whether through good luck or perspicacity (or a little of both), he finds himself facing this dilemma at a time when other major miners are letting project managers go in favor of cost cutting. So yet again, Mr. Glasenberg finds himself able to pick up attractive assets at even more attractive prices. We will be watching to see how the organization unfolds — will it be run to maximize LOM asset value, trade flow for the marketing division, or run in some hybrid fashion? Only time will tell. Time too will tell whether the senior traders at Glencore, on whom the strength of the reputation of its marketing division was built, will stick around as their post-IPO options vest — or whether this merger will help them monetize an exit strategy.

Looking farther ahead, the next logical move in Mr. Glasenberg's trajectory is, in our view, a bid for Anglo American. The synergies from the portfolios — particularly in coal — are attractive. Furthermore, "Glencore Xstrata Anglo" would form, in our view, a credible challenger to the "Big Three" of Vale, Rio and BHP, allowing the combined entity to more effectively compete for assets — and it would be in Glencore’s interests to be part of this transformation. But that is, in our view, a question for 2014-15.

48 To wit the April 21, 2013 story concerning Glencore's reported trading with Iranian supplier to nuclear program http://www.guardian.co.uk/business/2013/apr/21/glencore-trade-iran-supplier-nuclear. 49 Euro Metals & Mining: Not Enough Aristotle? The Fallacy of Growth & How Mining Companies Can Avoid Destroying Value.



AND LOVE THE ORE




Appendix 1: From Grizzly to De Niro

This appendix contains the analysis from the "What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities" chapter of this Blackbook and is divided into three parts.

Section 1 contains a summary of the equity scenario analysis — namely, how each of the equities in our coverage react under the "Grizzly," "Bear," "Bull," and "De Niro" scenarios relative to our current base scenario at the time the analysis was conducted (May 2013).

Section 2 offers the complete commodity prices applied across the four scenarios.

Section 3 shows the five-year historical spot price distribution of the key commodities used in these scenarios.

Please note that all analysis in this appendix, as in the corresponding chapter, was conducted in May 2013. All references to current share prices, spot prices and models are as of May 21, 2013.

Equity Scenario Analysis — Summary Exhibit 324 Vale Shows the Greatest Potential Upside and Widest Price Distribution Under Our

Scenarios (on the Back of Operational Gearing), While Glencore Shows the Greatest Downside and Least Upside Relative to Our Current Price Target (as a Result of Our Bullish View on Copper in the Medium Term)


TP / NPV per Share vs. Current Model

Current Model


Anglo American £ 23.00 6.51 17.03 26.84 36.22 -72% -26% 17% 57%

BHP Billiton £ 25.00 6.54 16.71 27.02 37.49 -74% -33% 8% 50%

Glen Xta Pro Forma £ 5.16 1.21 3.05 4.90 6.75 -77% -41% -5% 31%

Rio Tinto £ 45.75 15.23 37.73 59.99 81.69 -67% -18% 31% 79%

Vale BRL 29.25 9.73 34.38 58.97 83.54 -67% 18% 102% 186%



AND LOVE THE ORE

Exhibit 325 Rio Tinto Shows the Least Downside to Our Scenarios and Glencore the Greatest

Exhibit 326 As Soon as a More Positive Iron Ore Price Is Introduced, the Impact of Operational Gearing in Vale Is Immediately Apparent

Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis. Exhibit 327 Our Target Prices Are the Equivalent of the

"Bull" Scenario Even Though We Reach These Targets With Higher Short-Term Prices and Lower Long-Term Prices

Exhibit 328 It Is Difficult for Us to See Further Upside in Our Target Price for Glencore as It Already Factors in a Very Strong Copper Price


-67% -67%

-72%

-74%

-77%

Rio Tinto Vale Anglo American

BHP Billiton

Glen Xta Pro Forma

TP vs. Current Model — Grizzly

18%

-18%

-26%

-33%

-41%


BHP Billiton

Glen Xta Pro Forma

TP vs. Current Model — Bear

102%

31%

17%8%

-5%


BHP Billiton

Glen Xta Pro Forma

TP vs. Current Model — Bull

186%

79%

57%50%

31%


BHP Billiton

Glen Xta Pro Forma

TP vs. Current Model — De Niro




Exhibit 329 Rio Tinto Scenario Summary vs. Consensus

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 330 BHP Billiton Scenario Summary vs. Consensus

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 331 Anglo American Scenario Summary vs. Consensus


vs. Consensus

Rio TintoCurrent Model

Consensus Grizzly Bear Bull De Niro Grizzly Bear Bull De Niro

Price/Potential Upside £ 45.75 - 15.23 37.73 59.99 81.69 -49% 27% 103% 176%

Revenue 2012A US$m 50,967 50,967 50,967 50,967 50,967 50,967 - - - -

Revenue 2013E US$m 63,152 55,586 41,893 49,005 56,118 63,230 -25% -12% 1% 14%

Revenue 2014E US$m 79,156 61,445 46,879 55,078 63,278 71,477 -24% -10% 3% 16%

EBITDA 2012A US$m 20,381 20,381 20,381 20,381 20,381 20,381 - - - -

EBITDA 2013E US$m 30,054 21,007 8,102 15,491 22,879 30,267 -61% -26% 9% 44%

EBITDA 2014E US$m 41,048 23,251 7,887 16,344 24,802 33,259 -66% -30% 7% 43%

EPS 2012A US$m -1.62 -1.62 -1.62 -1.62 -1.62 -1.62 - - - -

EPS 2013E US$m 7.42 5.72 -0.32 2.00 4.33 6.67 -106% -65% -24% 17%

EPS 2014E US$m 10.41 6.62 -0.78 1.87 4.54 7.22 -112% -72% -31% 9%

vs. Consensus

BHP Billiton (June Year-End)Current Model


Price/Potential Upside £ 25.00 - 6.54 16.71 27.02 37.49 -64% -8% 49% 106%

Revenue 2012A US$m 72,226 72,226 72,226 72,226 72,226 72,226 - - - -

Revenue 2013E US$m 70,497 67,569 59,492 64,371 69,250 74,136 -12% -5% 2% 10%

Revenue 2014E US$m 93,105 73,914 58,919 69,648 80,376 91,119 -20% -6% 9% 23%

EBITDA 2012A US$m 30,324 30,324 30,324 30,324 30,324 30,324 - - - -

EBITDA 2013E US$m 28,861 28,727 17,932 22,776 27,620 32,471 -38% -21% -4% 13%

EBITDA 2014E US$m 50,886 33,552 16,925 27,581 38,238 48,909 -50% -18% 14% 46%

EPS 2012A US$m 2.93 2.93 2.93 2.93 2.93 2.93 - - - -

EPS 2013E US$m 3.20 2.53 1.74 2.35 2.96 3.57 -31% -7% 17% 41%

EPS 2014E US$m 5.64 3.05 1.03 2.38 3.73 5.08 -66% -22% 22% 67%

vs. Consensus

Anglo American Current Model


Price/Potential Upside £ 23.00 - 6.51 17.03 26.84 36.22 -59% 6% 67% 126%

Revenue 2012A US$m 28,761 28,761 28,761 28,761 28,761 28,761 - - - -

Revenue 2013E US$m 35,841 34,520 26,766 30,303 33,839 37,376 -22% -12% -2% 8%

Revenue 2014E US$m 41,500 36,177 27,806 31,526 35,247 38,968 -23% -13% -3% 8%

EBITDA 2012A US$m 8,815 8,815 8,815 8,815 8,815 8,815 - - - -

EBITDA 2013E US$m 13,412 9,657 4,288 7,810 11,333 14,855 -56% -19% 17% 54%

EBITDA 2014E US$m 17,136 10,177 3,353 7,058 10,762 14,466 -67% -31% 6% 42%

EPS 2012A US$m 2.22 2.22 2.22 2.22 2.22 2.22 - - - -

EPS 2013E US$m 2.86 2.17 -0.17 1.03 2.23 3.43 -108% -52% 3% 58%

EPS 2014E US$m 4.08 2.61 -0.66 0.63 1.91 3.18 -125% -76% -27% 22%



AND LOVE THE ORE

Exhibit 332 Glencore Xstrata Pro Forma Scenario Summary vs. Consensus

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 333 Vale Scenario Summary vs. Consensus


vs. Consensus

Glencore Xstrata Pro FormaCurrent Model


Price £ 5.16 - 1.21 3.05 4.90 6.75 -77% -41% -5% 31%

Revenue 2012 US$m 229,505 229,505 229,595 229,595 229,595 229,595 - - - -

Revenue 2013E US$m 251,245 244,469 203,872 228,386 252,901 277,415 -17% -7% 3% 13%

Revenue 2014E US$m 283,428 257,098 211,572 237,141 262,709 288,278 -18% -8% 2% 12%

EBITDA 2012 US$m 13,784 13,784 13,858 13,858 13,858 13,858 - - - -

EBITDA 2013E US$m 22,217 13,395 10,271 15,466 20,665 25,866 -23% 15% 54% 93%

EBITDA 2014E US$m 29,264 15,592 10,320 16,041 21,767 27,496 -34% 3% 40% 76%

EPS 2012 US$m 0.43 0.43 0.43 0.43 0.43 0.43 - - - -

EPS 2013E US$m 0.74 1.38 0.08 0.37 0.66 0.96 -94% -73% -52% -31%

EPS 2014E US$m 1.11 1.64 0.05 0.37 0.70 1.03 -97% -77% -57% -37%

vs. Consensus

ValeCurrent Model


Price/Potential Upside BRL 29.25 - 9.73 34.38 58.97 83.54 -72% 0% 72% 144%

Revenue 2012A US$m 47,694 47,694 47,694 47,694 47,694 47,694 - - - -

Revenue 2013E US$m 59,716 48,476 42,108 48,165 54,222 60,280 -13% -1% 12% 24%

Revenue 2014E US$m 74,202 50,580 41,150 50,065 58,980 67,895 -19% -1% 17% 34%

EBITDA 2012A US$m 18,133 18,133 18,133 18,133 18,133 18,133 - - - -

EBITDA 2013E US$m 31,950 21,994 14,342 20,399 26,456 32,513 -35% -7% 20% 48%

EBITDA 2014E US$m 43,806 22,151 10,753 19,668 28,583 37,499 -51% -11% 29% 69%

EPS 2012A US$m 1.07 1.07 1.07 1.07 1.07 1.07 - - - -

EPS 2013E US$m 3.73 2.27 1.11 2.02 2.95 3.90 -51% -11% 30% 71%

EPS 2014E US$m 5.48 2.19 0.45 1.80 3.17 4.59 -80% -18% 45% 110%




Exhibit 334 Both the "Grizzly"... Exhibit 335 ...And "Bear" Scenarios Show Revenue Below Consensus Expectations

Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis. Exhibit 336 Over the Next Two Years What We Call the

"Bull" Scenario Is Closest to Consensus Expectation

Exhibit 337 While Under the Strongest Price Scenario, There Is Upside to Consensus


-12%-13%

-17%

-22%-25%

-20%-19%

-18%

-23% -24%

BHP Billiton

Vale Glen Xta Pro Forma

Anglo American

Rio Tinto

Revenue vs. Consensus — Grizzly

2013E 2014E

-1%

-5%

-7%

-12% -12%

-1%

-6%

-8%

-10%

-13%

Vale BHP Billiton

Glen Xta Pro Forma

Rio Tinto Anglo American

Revenue vs. Consensus — Bear

2013E 2014E

12%

3%2%

1%

-2%

17%

2%

9%

3%

-3%

Vale Glen Xta Pro Forma

BHP Billiton

Rio Tinto Anglo American

Revenue vs. Consensus — Bull

2013E 2014E

24%

14% 13%

10%8%

34%

16%

12%

23%

8%

Vale Rio Tinto Glen Xta Pro Forma

BHP Billiton

Anglo American

Revenue vs. Consensus — De Niro

2013E 2014E



AND LOVE THE ORE

Exhibit 338 The Same Feature Is Seen at the EBITDA Line...

Exhibit 339 ...The Lack of Coverage of Glencore Makes Comparison More Difficult

Source: Corporate reports and Bernstein estimates and analysis. Source: Corporate reports and Bernstein estimates and analysis. Exhibit 340 Again, the "Bull" Scenario Appears Closest

to Consensus

Exhibit 341 Considerable Upside Under Stronger Commodity Scenarios Emphasizes the Degree of Operationally Geared Exposure the Miners Offer to Commodity Prices


-23%

-35%-38%

-56%-61%

-34%

-51% -50%

-67% -66%

Glen Xta Pro Forma

Vale BHP Billiton

Anglo American

Rio Tinto

EBITDA vs. Consensus — Grizzly

2013E 2014E

15%

-7%

-19% -21%

-26%

3%

-11%

-31%

-18%

-30%

Glen Xta Pro Forma

Vale Anglo American

BHP Billiton

Rio Tinto

EBITDA vs. Consensus — Bear

2013E 2014E

54%

20%17%

9%

-4%

40%

29%

6% 7%

14%

Glen Xta Pro Forma

Vale Anglo American

Rio Tinto BHP Billiton

EBITDA vs. Consensus — Bull

2013E 2014E

93%

54%48%

44%

13%

76%

42%

69%

43% 46%

Glen Xta Pro Forma

Anglo American


EBITDA vs. Consensus — De Niro

2013E 2014E




Rio Tinto Rio Tinto is primarily an iron ore business differentiated from its peers BHP Billiton and Vale by lack of petroleum on one side and by its high-quality (and growing) copper business on the other. Only under the most bullish price scenarios does aluminum offer diversification for the group on any measure other than simple revenue. Rio Tinto is our top pick, and, as this scenario analysis shows, the company offers the most attractive balance between upside potential and downside risk.

Exhibit 342 Rio Tinto Scenario Analysis Summary

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 343 Rio Tinto EV by Division


Rio Tinto Implied Share Price (GBP) vs. Current Model

Current Model 45.75 -

Grizzly 15.23 -67%

Bear 37.73 -18%

Bull 59.99 31%

De Niro 81.69 79%

174,957

77,886

150,387

221,759

291,325

(50,000)

-

50,000

100,000

150,000

200,000

250,000

300,000

350,000

Current Model Grizzly Bear Bull De Niro

US

$m

Rio Tinto EV

Iron Ore Aluminum Copper Energy Diamonds & Minerals Other Total



AND LOVE THE ORE

Exhibit 344 Rio Tinto — Current Model — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 345 Rio Tinto — Current Model — Valuation


Pro Forma Financials

Dec year-end 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A 2013 E 2014 E 2015 E

Income Statement

Consolidated revenue US$m 22,465 29,700 54,264 41,825 56,576 60,537 50,967 63,152 79,156 72,878

Operating profit US$m 10,352 10,155 11,233 8,292 20,795 14,052 (2,576) 21,070 30,268 20,651

PAT US$m 7,867 7,746 4,609 5,335 15,184 6,765 (3,004) 14,663 20,804 14,267

Minority interests US$m (429) (434) (933) (463) (860) (939) 14 (958) (1,582) (1,343)

Reported earnings US$m 7,438 7,312 3,676 4,872 14,324 5,826 (2,990) 13,706 19,222 12,924

Reported EPS, USD/sh USD/share 5.58 5.69 2.86 2.76 7.30 3.03 -1.62 7.42 10.41 7.00

Balance Sheet

Total assets US$m 34,494 101,391 89,616 97,236 112,773 119,545 117,573 126,444 140,674 149,858

Total liabilities US$m (15,109) (75,067) (67,155) (51,311) (48,261) (60,337) (59,552) (57,470) (55,988) (55,454)

Net assets US$m 19,385 26,324 22,461 45,925 64,512 59,208 58,021 68,974 84,686 94,403

Total equity US$m 19,385 26,324 22,461 45,925 64,512 59,208 58,021 68,974 84,686 94,403

Net debt/(Net Cash) US$m 2,775 45,182 38,577 18,861 4,071 8,807 19,412 17,829 6,220 (3,729)

Cash Flow Statement

EBITDA 0 12,566 13,611 23,870 14,471 26,639 29,586 20,381 30,054 41,048 31,540

Operating cash flow US$m 11,196 12,569 20,668 13,834 23,530 27,390 16,548 23,558 35,551 28,645

Net cash flow from operations US$m 8,076 8,491 14,883 9,212 18,277 20,032 9,466 16,584 24,896 21,308

Net cash flow from investing US$m (4,362) (42,742) (6,181) (3,357) (1,711) (16,838) (18,174) (12,249) (9,775) (8,152)

Net cash flow from financing US$m (5,389) 35,097 (9,108) (2,463) (10,610) (3,426) 6,325 (3,148) (6,202) (4,234)

Closing cash & equivalents US$m 736 1,645 1,181 4,233 9,948 9,670 7,082 8,269 17,188 26,110

FCF Per Share US$/share 3.36 3.23 6.38 3.04 7.57 4.61 -3.64 3.22 9.63 8.28

Valuation Summary

Price Target/History £ 22.45 44.36 11.36 33.70 44.87 31.25 35.12 45.75 45.75 45.75

Price Target/History $ 43.97 88.13 16.72 53.85 69.39 48.58 56.63 72.50 72.50 72.50

MCAP US$m 57,574 113,316 21,456 94,963 136,070 92,178 104,574 133,865 133,865 133,865

EV US$m 63,669 165,224 65,479 122,848 148,311 115,842 124,475 174,957 174,957 174,957

EV/EBITDA - 5.1 12.1 2.7 8.5 5.6 3.9 6.1 5.8 4.3 5.5

PE - 7.9 15.5 5.8 19.5 9.5 16.0 -35.0 9.8 7.0 10.4


Iron Ore 69,188 15,979 85,167 46.12 29.11 20,945 4.1

Aluminum 15,664 6,259 21,923 11.87 7.49 2,168 10.1

Copper 38,478 7,837 46,315 25.08 15.83 3,318 14.0

Energy 13,894 2,553 16,447 8.91 5.62 2,693 6.1


Other (3,043) - (3,043) (1.65) (1.04) (360) 8.5


Less Minorities (20,530) (11.12) (7.02)



Equity Value 133,865 72.50 45.75



Exchange Rate 1.58





Exhibit 346 Rio Tinto — Grizzly — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 347 Rio Tinto — Grizzly — Valuation




Income Statement


Operating profit US$m 10,352 10,155 11,233 8,292 20,795 14,052 (2,576) 266 (1,079) 79

PAT US$m 7,867 7,746 4,609 5,335 15,184 6,765 (3,004) (940) (1,948) (1,201)

Minority interests US$m (429) (434) (933) (463) (860) (939) 14 351 504 285

Reported earnings US$m 7,438 7,312 3,676 4,872 14,324 5,826 (2,990) (588) (1,445) (916)

Reported EPS USD/share 5.58 5.69 2.86 2.76 7.30 3.03 -1.62 -0.32 -0.78 -0.50

Balance Sheet

Total assets US$m 34,494 101,391 89,616 97,236 112,773 119,545 117,573 115,634 116,112 116,179


Net assets US$m 19,385 26,324 22,461 45,925 64,512 59,208 58,021 56,204 54,910 54,281

Total equity US$m 19,385 26,324 22,461 45,925 64,512 59,208 58,021 56,204 54,910 54,281

Net debt/(Net cash) US$m 2,775 45,182 38,577 18,861 4,071 8,807 19,412 25,896 28,276 29,094

Cash Flow Statement

EBITDA US$m 12,566 13,611 23,870 14,471 26,639 29,586 20,381 8,102 7,887 9,311




Net cash flow from financing US$m (5,389) 35,097 (9,108) (2,463) (10,610) (3,426) 6,325 393 745 492

Closing cash & equivalents US$m 736 1,645 1,181 4,233 9,948 9,670 7,082 2,219 434 (179)

FCF per share US$/share 3.36 3.23 6.38 3.04 7.57 4.61 -3.64 -2.66 -0.89 0.04

Valuation Summary



MCAP US$m 57,574 113,316 21,456 94,963 136,070 92,178 104,574 44,566 44,566 44,566

EV US$m 63,669 165,224 65,479 122,848 148,311 115,842 124,475 77,886 77,886 77,886

EV/EBITDA - 5.1 12.1 2.7 8.5 5.6 3.9 6.1 9.6 9.9 8.4

PE - 7.9 15.5 5.8 19.5 9.5 16.0 -35.0 -75.8 -30.8 -48.6


Iron Ore 40,216 14,407 54,623 29.58 18.67 6,225 8.8

Aluminum (6,068) - (6,068) (3.29) (2.07) (619) 9.8

Copper 16,066 6,231 22,297 12.08 7.62 1,260 17.7

Energy 1,435 1,099 2,535 1.37 .87 452 5.6


Other (3,110) - (3,110) (1.68) (1.06) (360) 8.6


Less Minorities (9,660) (5.23) (3.30)



Equity Value 44,566 24.14 15.23



Exchange Rate 1.58




AND LOVE THE ORE

Exhibit 348 Rio Tinto — Bear — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 349 Rio Tinto — Bear — Valuation




Income Statement



PAT US$m 7,867 7,746 4,609 5,335 15,184 6,765 (3,004) 3,802 3,513 5,024

Minority interests US$m (429) (434) (933) (463) (860) (939) 14 (115) (66) (364)


Reported EPS USD/share 5.58 5.69 2.86 2.76 7.30 3.03 -1.62 2.00 1.87 2.52

Balance Sheet

Total assets US$m 34,494 101,391 89,616 97,236 112,773 119,545 117,573 118,872 122,477 126,105


Net assets US$m 19,385 26,324 22,461 45,925 64,512 59,208 58,021 59,981 62,651 66,546

Total equity US$m 19,385 26,324 22,461 45,925 64,512 59,208 58,021 59,981 62,651 66,546


Cash Flow Statement

EBITDA US$m 12,566 13,611 23,870 14,471 26,639 29,586 20,381 15,491 16,344 18,876




Net cash flow from financing US$m (5,389) 35,097 (9,108) (2,463) (10,610) (3,426) 6,325 (673) (1,026) (1,532)


FCF per share US$/share 3.36 3.23 6.38 3.04 7.57 4.61 -3.64 -0.91 1.70 2.96

Valuation Summary



MCAP US$m 57,574 113,316 21,456 94,963 136,070 92,178 104,574 110,378 110,378 110,378

EV US$m 63,669 165,224 65,479 122,848 148,311 115,842 124,475 150,387 150,387 150,387

EV/EBITDA - 5.1 12.1 2.7 8.5 5.6 3.9 6.1 9.7 9.2 8.0

PE - 7.9 15.5 5.8 19.5 9.5 16.0 -35.0 29.9 32.0 23.7


Iron Ore 76,810 23,393 100,203 54.27 34.25 10,716 9.4

Aluminum 723 1,058 1,781 .96 .61 535 3.3

Copper 25,248 8,178 33,426 18.10 11.43 2,114 15.8

Energy 7,318 2,818 10,136 5.49 3.46 1,284 7.9


Other (3,097) - (3,097) (1.68) (1.06) (360) 8.6


Less Minorities (17,202) (9.32) (5.88)



Equity Value 110,378 59.78 37.73



Exchange Rate 1.58





Exhibit 350 Rio Tinto — Bull — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 351 Rio Tinto — Bull — Valuation




Income Statement



PAT US$m 7,867 7,746 4,609 5,335 15,184 6,765 (3,004) 8,543 8,975 11,250

Minority interests US$m (429) (434) (933) (463) (860) (939) 14 (539) (584) (963)



Balance Sheet

Total assets US$m 34,494 101,391 89,616 97,236 112,773 119,545 117,573 122,137 128,903 136,121


Net assets US$m 19,385 26,324 22,461 45,925 64,512 59,208 58,021 63,795 70,472 78,933

Total equity US$m 19,385 26,324 22,461 45,925 64,512 59,208 58,021 63,795 70,472 78,933


Cash Flow Statement

EBITDA US$m 12,566 13,611 23,870 14,471 26,639 29,586 20,381 22,879 24,802 28,441






FCF per share US$/share 3.36 3.23 6.38 3.04 7.57 4.61 -3.64 0.84 4.30 5.89

Valuation Summary



MCAP US$m 57,574 113,316 21,456 94,963 136,070 92,178 104,574 175,496 175,496 175,496

EV US$m 63,669 165,224 65,479 122,848 148,311 115,842 124,475 221,759 221,759 221,759

EV/EBITDA - 5.1 12.1 2.7 8.5 5.6 3.9 6.1 9.7 8.9 7.8

PE - 7.9 15.5 5.8 19.5 9.5 16.0 -35.0 21.9 20.9 17.1


Iron Ore 112,270 31,795 144,065 78.02 49.24 15,208 9.5

Aluminum 7,437 3,375 10,812 5.86 3.70 1,690 6.4

Copper 34,262 9,984 44,246 23.96 15.12 2,968 14.9

Energy 13,029 4,437 17,465 9.46 5.97 2,117 8.3


Other (3,079) - (3,079) (1.67) (1.05) (360) 8.6


Less Minorities (24,330) (13.18) (8.32)



Equity Value 175,496 95.04 59.99



Exchange Rate 1.58




AND LOVE THE ORE

Exhibit 352 Rio Tinto — De Niro — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 353 Rio Tinto — De Niro — Valuation




Income Statement



PAT US$m 7,867 7,746 4,609 5,335 15,184 6,765 (3,004) 13,284 14,438 17,477

Minority interests US$m (429) (434) (933) (463) (860) (939) 14 (962) (1,098) (1,558)



Balance Sheet

Total assets US$m 34,494 101,391 89,616 97,236 112,773 119,545 117,573 125,403 135,331 147,076


Net assets US$m 19,385 26,324 22,461 45,925 64,512 59,208 58,021 67,609 78,297 91,326

Total equity US$m 19,385 26,324 22,461 45,925 64,512 59,208 58,021 67,609 78,297 91,326

Net debt/(Net cash) US$m 2,775 45,182 38,577 18,861 4,071 8,807 19,412 19,004 11,178 251

Cash Flow Statement

EBITDA US$m 12,566 13,611 23,870 14,471 26,639 29,586 20,381 30,267 33,259 38,007






FCF per share US$/share 3.36 3.23 6.38 3.04 7.57 4.61 -3.64 2.58 6.89 8.82

Valuation Summary



MCAP US$m 57,574 113,316 21,456 94,963 136,070 92,178 104,574 238,988 238,988 238,988

EV US$m 63,669 165,224 65,479 122,848 148,311 115,842 124,475 291,325 291,325 291,325

EV/EBITDA - 5.1 12.1 2.7 8.5 5.6 3.9 6.1 9.6 8.8 7.7

PE - 7.9 15.5 5.8 19.5 9.5 16.0 -35.0 19.4 17.9 15.0


Iron Ore 147,127 39,709 186,836 101.18 63.86 19,700 9.5

Aluminum 14,054 5,545 19,599 10.61 6.70 2,844 6.9

Copper 43,122 11,672 54,795 29.67 18.73 3,822 14.3

Energy 18,651 5,952 24,603 13.32 8.41 2,949 8.3


Other (3,059) - (3,059) (1.66) (1.05) (360) 8.5


Less Minorities (31,279) (16.94) (10.69)



Equity Value 238,988 129.43 81.69



Exchange Rate 1.58





BHP Billiton BHP Billiton is the highest quality stock in our coverage, and has the second most diversified portfolio after Anglo American. But where Anglo has the drag of platinum, BHP has a petroleum kicker (run at 77% EBITDA margins in 2011 and 73% in 2012). BHP fares second worst in the Grizzly scenario (after Glencore) but only third best in the De Niro scenario (the inverse of Anglo’s third worst in Grizzly and second best in De Niro). The presence of petroleum means that, uniquely for BHP, there is a third key explanatory variable (besides copper and iron ore).

Exhibit 354 BHP Billiton Scenario Analysis Summary

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 355 BHP Billiton EV by Division


BHP Billiton Implied Share Price (GBP) vs. Current Model


Grizzly 6.54 -74%

Bear 16.71 -33%

Bull 27.02 8%

De Niro 37.49 50%

251,343

94,981

182,557

270,449

359,245

(50,000)

-

50,000

100,000

150,000

200,000

250,000

300,000

350,000

400,000


US

$m

BHP Billiton EV

Petroleum Aluminium Base MetalsDiamonds & Specialty Products Stainless Steel Materials Iron OreMetallurgical Coal Energy Coal Manganese



AND LOVE THE ORE

Exhibit 356 BHP Billiton — Current Model — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 357 BHP Billiton — Current Model — Valuation



June year-end 2006 A 2007 A 2008 A 2009 A 2010 A 2011 A 2012 A 2013 E 2014 E 2015 E

Income Statement

Group revenue US$m 39,099 47,473 59,473 50,211 52,798 71,739 72,226 70,497 93,105 101,053

Operating profit US$m 14,671 18,401 24,145 12,160 20,030 31,816 23,916 24,835 43,813 45,665

PAT US$m 10,534 13,496 15,962 6,338 13,008 23,946 15,696 17,218 30,357 32,172

Minority interests US$m (84) (80) (572) (461) (287) (298) (115) (177) (361) (345)

Reported Earnings US$m 10,450 13,416 15,390 5,877 12,721 23,648 15,581 17,041 29,997 31,827

Reported EPS, USD/sh USD/share 1.76 2.39 2.77 1.06 2.29 4.44 2.93 3.20 5.64 5.98

Balance Sheet

Total assets US$m 48,516 58,168 76,008 78,770 88,852 102,891 129,273 138,378 157,766 179,545


Net assets US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 75,170 97,983 121,951

Total equity US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 75,170 97,983 121,951

Net debt/(Net cash) US$m 8,240 8,706 8,458 5,586 3,308 5,823 23,549 30,570 15,735 (4,393)

Cash Flow Statement

EBITDA 0 17,447 21,284 27,757 16,031 24,790 36,855 30,324 28,861 50,886 52,921




Net cash flow from financing US$m (5,412) (6,843) (6,999) (1,180) (5,307) (16,018) 2,509 326 (11,344) (10,840)


FCF Per Share US$/share 0.50 0.95 1.95 1.74 1.43 3.64 1.26 -0.18 4.49 5.52

Valuation Summary



MCAP US$m 113,795 156,214 206,507 123,855 154,816 155,346 182,747 210,242 210,242 210,242

EV US$m 122,950 165,852 222,640 139,155 161,616 163,126 207,645 251,343 251,343 251,343

EV/EBITDA - 7.0 7.8 8.0 8.7 6.5 4.4 6.8 8.7 4.9 4.7

PE - 10.9 11.6 13.4 21.1 12.2 6.6 11.7 12.3 7.0 6.6


Petroleum 91,182 - 91,182 17.14 10.82 12,170 7.5

Aluminum 5,580 942 6,522 1.23 .77 549 11.9

Base Metals 43,619 8,442 52,061 9.78 6.18 6,658 7.8


Stainless Steel Materials 47 - 47 .01 .01 391 .1

Iron Ore 54,750 11,663 66,413 12.48 7.88 15,752 4.2

Manganese 3,743 56 3,799 .71 .45 831 4.6


Energy Coal 8,090 1,310 9,399 1.77 1.11 1,664 5.6



Less Central Costs (8,032) (1.51) (.95)


Equity Value 210,242 39.51 25.00



Exchange Rate 1.58





Exhibit 358 BHP Billiton — Grizzly — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 359 BHP Billiton — Grizzly — Valuation




Income Statement



PAT US$m 10,534 13,496 15,962 6,338 13,008 23,946 15,696 9,312 5,495 4,707

Minority interests US$m (84) (80) (572) (461) (287) (298) (115) (65) (12) 23

Reported earnings US$m 10,450 13,416 15,390 5,877 12,721 23,648 15,581 9,247 5,482 4,730

Reported EPS USD/share 1.76 2.39 2.77 1.06 2.29 4.44 2.93 1.74 1.03 0.89

Balance Sheet

Total assets US$m 48,516 58,168 76,008 78,770 88,852 102,891 129,273 131,685 136,149 140,178


Net assets US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 67,376 71,282 74,872

Total equity US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 67,376 71,282 74,872


Cash Flow Statement

EBITDA US$m 17,447 21,284 27,757 16,031 24,790 36,855 30,324 17,932 16,925 15,887




Net cash flow from financing US$m (5,412) (6,843) (6,999) (1,180) (5,307) (16,018) 2,509 1,573 (1,388) (1,139)


FCF per share US$/share 0.50 0.95 1.95 1.74 1.43 3.64 1.26 -1.03 0.47 0.55

Valuation Summary



MCAP US$m 113,795 156,214 206,507 123,855 154,816 155,346 182,747 55,132 55,132 55,132

EV US$m 122,950 165,852 222,640 139,155 161,616 163,126 207,645 94,981 94,981 94,981

EV/EBITDA - 7.0 7.8 8.0 8.7 6.5 4.4 6.8 5.3 5.6 6.0

PE - 10.9 11.6 13.4 21.1 12.2 6.6 11.7 6.0 10.1 11.7


Petroleum 33,627 - 33,627 6.32 3.99 8,301 4.1

Aluminum (3,929) - (3,929) (.74) (.47) (551) 7.1

Base Metals 14,280 4,787 19,067 3.58 2.26 2,731 7.0


Stainless Steel Materials (1,556) - (1,556) (.29) (.18) (71) 22.0

Iron Ore 31,758 9,792 41,550 7.81 4.93 4,391 9.5

Manganese 1,157 37 1,194 .22 .14 218 5.5

Metallurgical Coal 3,538 2,135 5,673 1.07 .67 650 8.7

Energy Coal 2,365 835 3,200 .60 .38 565 5.7


Less Minorities (605) (.11) (.07)

Less Central Costs (4,246) (.80) (.50)


Equity Value 55,132 10.36 6.54



Exchange Rate 1.58




AND LOVE THE ORE

Exhibit 360 BHP Billiton — Bear — Pro Forma Financials




Income Statement



PAT US$m 10,534 13,496 15,962 6,338 13,008 23,946 15,696 12,606 12,783 12,853




Balance Sheet

Total assets US$m 48,516 58,168 76,008 78,770 88,852 102,891 129,273 134,493 142,870 151,650


Net assets US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 70,617 79,628 89,452

Total equity US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 70,617 79,628 89,452


Cash Flow Statement

EBITDA US$m 17,447 21,284 27,757 16,031 24,790 36,855 30,324 22,776 27,581 27,688







Valuation Summary



MCAP US$m 113,795 156,214 206,507 123,855 154,816 155,346 182,747 140,900 140,900 140,900

EV US$m 122,950 165,852 222,640 139,155 161,616 163,126 207,645 182,557 182,557 182,557

EV/EBITDA - 7.0 7.8 8.0 8.7 6.5 4.4 6.8 8.0 6.6 6.6

PE - 10.9 11.6 13.4 21.1 12.2 6.6 11.7 11.3 11.1 11.0




Exhibit 361 BHP Billiton — Bear — Valuation




Income Statement



PAT US$m 10,534 13,496 15,962 6,338 13,008 23,946 15,696 12,606 12,783 12,853




Balance Sheet

Total assets US$m 48,516 58,168 76,008 78,770 88,852 102,891 129,273 134,493 142,870 151,650


Net assets US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 70,617 79,628 89,452

Total equity US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 70,617 79,628 89,452


Cash Flow Statement

EBITDA US$m 17,447 21,284 27,757 16,031 24,790 36,855 30,324 22,776 27,581 27,688







Valuation Summary



MCAP US$m 113,795 156,214 206,507 123,855 154,816 155,346 182,747 140,900 140,900 140,900

EV US$m 122,950 165,852 222,640 139,155 161,616 163,126 207,645 182,557 182,557 182,557

EV/EBITDA - 7.0 7.8 8.0 8.7 6.5 4.4 6.8 8.0 6.6 6.6

PE - 10.9 11.6 13.4 21.1 12.2 6.6 11.7 11.3 11.1 11.0



AND LOVE THE ORE

Exhibit 362 BHP Billiton — Bull — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 363 BHP Billiton — Bull — Valuation




Income Statement



PAT US$m 10,534 13,496 15,962 6,338 13,008 23,946 15,696 15,900 20,072 20,999




Balance Sheet

Total assets US$m 48,516 58,168 76,008 78,770 88,852 102,891 129,273 137,301 149,592 163,123


Net assets US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 73,858 87,973 104,032

Total equity US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 73,858 87,973 104,032


Cash Flow Statement

EBITDA US$m 17,447 21,284 27,757 16,031 24,790 36,855 30,324 27,620 38,238 39,489







Valuation Summary



MCAP US$m 113,795 156,214 206,507 123,855 154,816 155,346 182,747 227,777 227,777 227,777

EV US$m 122,950 165,852 222,640 139,155 161,616 163,126 207,645 270,449 270,449 270,449

EV/EBITDA - 7.0 7.8 8.0 8.7 6.5 4.4 6.8 9.8 7.1 6.8

PE - 10.9 11.6 13.4 21.1 12.2 6.6 11.7 14.5 11.5 11.0


Petroleum 73,891 - 73,891 13.89 8.76 13,170 5.6

Aluminum 1,889 73 1,962 .37 .23 354 5.5

Base Metals 35,621 10,746 46,367 8.71 5.50 6,008 7.7


Stainless Steel Materials 1,654 487 2,141 .40 .25 489 4.4

Iron Ore 89,025 22,698 111,723 21.00 13.25 11,348 9.8

Manganese 4,059 88 4,147 .78 .49 759 5.5


Energy Coal 7,880 2,498 10,378 1.95 1.23 1,422 7.3





Equity Value 227,777 42.81 27.02



Exchange Rate 1.58





Exhibit 364 BHP Billiton — De Niro — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 365 BHP Billiton — De Niro — Valuation




Income Statement



PAT US$m 10,534 13,496 15,962 6,338 13,008 23,946 15,696 19,199 27,370 29,141




Balance Sheet

Total assets US$m 48,516 58,168 76,008 78,770 88,852 102,891 129,273 140,113 156,322 176,258


Net assets US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 77,104 96,329 118,618

Total equity US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 77,104 96,329 118,618

Net debt/(Net cash) US$m 8,240 8,706 8,458 5,586 3,308 5,823 23,549 29,772 16,634 (882)

Cash Flow Statement

EBITDA US$m 17,447 21,284 27,757 16,031 24,790 36,855 30,324 32,471 48,909 51,290







Valuation Summary



MCAP US$m 113,795 156,214 206,507 123,855 154,816 155,346 182,747 316,027 316,027 316,027

EV US$m 122,950 165,852 222,640 139,155 161,616 163,126 207,645 359,245 359,245 359,245

EV/EBITDA - 7.0 7.8 8.0 8.7 6.5 4.4 6.8 11.1 7.3 7.0

PE - 10.9 11.6 13.4 21.1 12.2 6.6 11.7 16.7 11.7 11.0


Petroleum 94,043 - 94,043 17.67 11.16 15,605 6.0

Aluminum 4,781 688 5,469 1.03 .65 806 6.8

Base Metals 46,296 13,731 60,028 11.28 7.12 7,647 7.9


Stainless Steel Materials 3,252 843 4,095 .77 .49 769 5.3

Iron Ore 117,657 29,160 146,817 27.59 17.41 14,826 9.9

Manganese 5,509 113 5,623 1.06 .67 1,029 5.5


Energy Coal 10,899 3,446 14,346 2.70 1.70 1,873 7.7





Equity Value 316,027 59.39 37.49



Exchange Rate 1.58




AND LOVE THE ORE

Anglo American Anglo American has the most commodity-diversified portfolio in our coverage and the second-highest country risk (after Glencore), thanks to its South African exposure. Where BHP’s analysis confirms a third explanatory variable (oil), Anglo’s is impacted by the interlinked risks of platinum and South African country risk. The company fares third worst in the Grizzly scenario and second best in De Niro. We will look with interest to see how the Anglo turnaround story under new CEO Mark Cutifani plays out and how resultant changes in exposure may alter the potential upside/downside of this analysis.

Exhibit 366 Anglo American Scenario Analysis Summary

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 367 Anglo American EV by Division


Anglo American Implied Share Price (GBP) vs. Current Model


Grizzly 6.51 -72%

Bear 17.03 -26%

Bull 26.84 17%

De Niro 36.22 57%

71,480

30,791

57,712

82,848

106,894

(20,000)

-

20,000

40,000

60,000

80,000

100,000

120,000


US

$m

Anglo American EV

Platinum Diamonds Copper Nickel

Iron Ore - South Africa Iron Ore - Brazil Manganese Metallurgical Coal

Thermal Coal Other - Non Mining Assets




Exhibit 368 Anglo American — Current Model — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 369 Anglo American — Current Model — Valuation




Income Statement


Operating profit (underlying) US$m 9,324 9,158 9,284 4,695 9,353 10,646 5,898 8,835 11,979 9,948

PAT (underlying) US$m 6,396 6,345 6,287 3,117 6,410 7,885 4,122 5,653 7,866 6,542

Minority interests US$m (925) (902) (1,050) (548) (1,434) (1,765) (1,283) (2,004) (2,650) (1,767)

Underlying Earnings US$m 5,471 5,443 5,237 2,569 4,976 6,120 2,839 3,649 5,216 4,775

Underlying EPS, USD/sh USD/share 3.74 4.47 4.36 2.14 4.13 5.05 2.22 2.86 4.08 3.74

Balance Sheet

Total assets US$m 46,483 44,762 49,738 56,308 66,656 72,442 79,369 82,062 86,671 90,367


Net assets US$m 27,127 24,330 21,756 28,069 37,971 43,189 43,787 47,135 52,306 56,730

Total equity US$m 27,127 24,330 21,756 28,069 37,971 43,189 43,787 47,135 52,306 56,730

Net debt/(Net Cash) US$m 3,051 4,782 11,348 11,328 7,443 1,374 8,828 5,255 2,511 (3,234)

Cash Flow Statement

EBITDA 0 12,197 11,171 11,790 6,930 11,983 13,463 8,815 13,412 17,136 16,058




Net cash flow from financing US$m (6,675) (4,969) 3,542 (1,619) (2,400) 1,474 1,950 (5,357) (5,200) (4,298)

Closing cash & equivalents US$m 3,004 3,129 2,771 3,269 6,401 11,732 9,094 10,459 11,453 15,687

FCF Per Share US$/share 3.16 2.74 2.38 -0.43 2.03 2.61 -0.61 3.34 4.67 6.50

Valuation Summary



MCAP US$m 73,390 74,937 25,099 52,066 62,214 44,788 39,038 46,650 46,650 46,650

EV US$m 88,850 92,138 41,479 74,501 87,586 59,078 65,509 71,480 71,480 71,480

EV/EBITDA - 7.3 8.2 3.5 10.8 7.3 4.4 7.4 5.3 4.2 4.5

PE - 6.8 6.9 3.3 12.7 8.1 4.7 8.5 8.1 5.6 6.2


Platinum 5,699 1,190 6,889 5.39 3.40 1,447 4.8

Diamonds 5,368 - 5,368 4.20 2.65 1,437 3.7

Copper 21,244 3,084 24,328 19.04 12.01 2,927 8.3

Nickel 1,370 239 1,609 1.26 .79 200 8.0


Iron Ore — Brazil 1,455 1,301 2,756 2.16 1.36 227 12.2

Manganese 2,826 - 2,826 2.21 1.40 615 4.6

Metallurgical Coal 6,563 683 7,247 5.67 3.58 1,399 5.2

Thermal Coal 6,010 234 6,244 4.89 3.08 1,320 4.7



Less Minorities (17,365) (13.59) (8.58)



Equity Value 46,650 36.50 23.00



Exchange Rate 1.58




AND LOVE THE ORE

Exhibit 370 Anglo American — Grizzly — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 371 Anglo American — Grizzly — Valuation




Income Statement


Operating profit (underlying) US$m 9,324 9,158 9,284 4,695 9,353 10,646 5,898 656 (240) (252)

PAT (underlying) US$m 6,396 6,345 6,287 3,117 6,410 7,885 4,122 125 (509) (567)

Minority interests US$m (925) (902) (1,050) (548) (1,434) (1,765) (1,283) (345) (330) (371)

Underlying earnings US$m 5,471 5,443 5,237 2,569 4,976 6,120 2,839 (221) (839) (938)

Underlying EPS USD/share 3.74 4.47 4.36 2.14 4.13 5.05 2.22 -0.17 -0.66 -0.73

Balance Sheet

Total assets US$m 46,483 44,762 49,738 56,308 66,656 72,442 79,369 78,833 79,678 78,664


Net assets US$m 27,127 24,330 21,756 28,069 37,971 43,189 43,787 43,009 42,360 41,743

Total equity US$m 27,127 24,330 21,756 28,069 37,971 43,189 43,787 43,009 42,360 41,743


Cash Flow Statement

EBITDA US$m 12,197 11,171 11,790 6,930 11,983 13,463 8,815 4,288 3,353 4,317




Net cash flow from financing US$m (6,675) (4,969) 3,542 (1,619) (2,400) 1,474 1,950 (2,560) (205) (1,735)


FCF per share US$/share 3.16 2.74 2.38 -0.43 2.03 2.61 -0.61 0.04 -1.30 1.13

Valuation Summary



MCAP US$m 73,390 74,937 25,099 52,066 62,214 44,788 39,038 13,184 13,184 13,184

EV US$m 88,850 92,138 41,479 74,501 87,586 59,078 65,509 30,791 30,791 30,791

EV/EBITDA - 7.3 8.2 3.5 10.8 7.3 4.4 7.4 7.2 9.2 7.1

PE - 6.8 6.9 3.3 12.7 8.1 4.7 8.5 -37.7 -9.9 -8.9


Platinum 1,578 986 2,563 2.01 1.27 415 6.2

Diamonds 3,368 - 3,368 2.64 1.66 660 5.1

Copper 9,608 2,174 11,782 9.22 5.82 1,492 7.9

Nickel 1,023 243 1,266 .99 .63 128 9.9

Iron Ore — South Africa 5,006 972 5,978 4.68 2.95 639 9.4

Iron Ore — Brazil 718 1,136 1,854 1.45 .92 83 22.4

Manganese 1,516 - 1,516 1.19 .75 299 5.1

Metallurgical Coal (2,844) - (2,844) (2.23) (1.40) (136) 20.9

Thermal Coal 2,558 197 2,754 2.16 1.36 603 4.6



Less Minorities (9,008) (7.05) (4.45)



Equity Value 13,184 10.32 6.51



Exchange Rate 1.58





Exhibit 372 Anglo American — Bear — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 373 Anglo American — Bear — Valuation




Income Statement




Minority interests US$m (925) (902) (1,050) (548) (1,434) (1,765) (1,283) (941) (962) (1,021)

Underlying earnings US$m 5,471 5,443 5,237 2,569 4,976 6,120 2,839 1,320 800 1,030

Underlying EPS USD/share 3.74 4.47 4.36 2.14 4.13 5.05 2.22 1.03 0.63 0.81

Balance Sheet

Total assets US$m 46,483 44,762 49,738 56,308 66,656 72,442 79,369 80,040 81,801 81,928


Net assets US$m 27,127 24,330 21,756 28,069 37,971 43,189 43,787 44,608 45,502 46,727

Total equity US$m 27,127 24,330 21,756 28,069 37,971 43,189 43,787 44,608 45,502 46,727


Cash Flow Statement

EBITDA US$m 12,197 11,171 11,790 6,930 11,983 13,463 8,815 7,810 7,058 8,508






FCF per share US$/share 3.16 2.74 2.38 -0.43 2.03 2.61 -0.61 1.37 0.42 3.03

Valuation Summary



MCAP US$m 73,390 74,937 25,099 52,066 62,214 44,788 39,038 34,477 34,477 34,477

EV US$m 88,850 92,138 41,479 74,501 87,586 59,078 65,509 57,712 57,712 57,712

EV/EBITDA - 7.3 8.2 3.5 10.8 7.3 4.4 7.4 7.4 8.2 6.8

PE - 6.8 6.9 3.3 12.7 8.1 4.7 8.5 16.5 27.2 21.1


Platinum 4,982 1,702 6,685 5.23 3.30 1,005 6.7

Diamonds 4,770 - 4,770 3.73 2.36 969 4.9

Copper 13,451 2,848 16,298 12.75 8.05 2,127 7.7

Nickel 1,337 294 1,632 1.28 .81 171 9.5


Iron Ore — Brazil 2,909 1,688 4,597 3.60 2.27 131 35.1

Manganese 2,257 - 2,257 1.77 1.11 438 5.1

Metallurgical Coal 1,455 513 1,967 1.54 .97 390 5.0

Thermal Coal 4,011 255 4,266 3.34 2.11 888 4.8



Less Minorities (14,813) (11.59) (7.32)



Equity Value 34,477 26.98 17.03



Exchange Rate 1.58




AND LOVE THE ORE

Exhibit 374 Anglo American — Bull — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 375 Anglo American — Bull — Valuation




Income Statement





Underlying earnings US$m 5,471 5,443 5,237 2,569 4,976 6,120 2,839 2,850 2,435 2,991


Balance Sheet

Total assets US$m 46,483 44,762 49,738 56,308 66,656 72,442 79,369 81,232 83,904 85,630


Net assets US$m 27,127 24,330 21,756 28,069 37,971 43,189 43,787 46,199 48,632 51,692

Total equity US$m 27,127 24,330 21,756 28,069 37,971 43,189 43,787 46,199 48,632 51,692

Net debt/(Net cash) US$m 3,051 4,782 11,348 11,328 7,443 1,374 8,828 5,735 5,149 946

Cash Flow Statement

EBITDA US$m 12,197 11,171 11,790 6,930 11,983 13,463 8,815 11,333 10,762 12,699







Valuation Summary



MCAP US$m 73,390 74,937 25,099 52,066 62,214 44,788 39,038 54,344 54,344 54,344

EV US$m 88,850 92,138 41,479 74,501 87,586 59,078 65,509 82,848 82,848 82,848

EV/EBITDA - 7.3 8.2 3.5 10.8 7.3 4.4 7.4 7.3 7.7 6.5

PE - 6.8 6.9 3.3 12.7 8.1 4.7 8.5 12.0 14.1 11.5


Platinum 8,298 2,340 10,638 8.32 5.25 1,595 6.7

Diamonds 6,142 - 6,142 4.81 3.03 1,277 4.8

Copper 17,213 3,453 20,666 16.17 10.21 2,762 7.5

Nickel 1,630 340 1,970 1.54 .97 215 9.2


Iron Ore — Brazil 5,000 2,183 7,183 5.62 3.55 179 40.1

Manganese 2,981 - 2,981 2.33 1.47 578 5.2

Metallurgical Coal 5,011 994 6,005 4.70 2.97 916 6.6

Thermal Coal 5,435 309 5,744 4.49 2.84 1,173 4.9



Less Minorities (20,449) (16.00) (10.10)



Equity Value 54,344 42.52 26.84



Exchange Rate 1.58





Exhibit 376 Anglo American — De Niro — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 377 Anglo American — De Niro — Valuation




Income Statement







Balance Sheet

Total assets US$m 46,483 44,762 49,738 56,308 66,656 72,442 79,369 82,417 85,995 90,141


Net assets US$m 27,127 24,330 21,756 28,069 37,971 43,189 43,787 47,787 51,755 56,644

Total equity US$m 27,127 24,330 21,756 28,069 37,971 43,189 43,787 47,787 51,755 56,644

Net debt/(Net cash) US$m 3,051 4,782 11,348 11,328 7,443 1,374 8,828 4,550 2,452 (3,469)

Cash Flow Statement

EBITDA US$m 12,197 11,171 11,790 6,930 11,983 13,463 8,815 14,855 14,466 16,889







Valuation Summary



MCAP US$m 73,390 74,937 25,099 52,066 62,214 44,788 39,038 73,350 73,350 73,350

EV US$m 88,850 92,138 41,479 74,501 87,586 59,078 65,509 106,894 106,894 106,894

EV/EBITDA - 7.3 8.2 3.5 10.8 7.3 4.4 7.4 7.2 7.4 6.3

PE - 6.8 6.9 3.3 12.7 8.1 4.7 8.5 10.6 11.4 9.4


Platinum 11,526 2,905 14,432 11.29 7.13 2,184 6.6

Diamonds 7,480 - 7,480 5.85 3.69 1,585 4.7

Copper 20,886 3,989 24,875 19.46 12.28 3,396 7.3

Nickel 1,914 381 2,295 1.80 1.13 258 8.9


Iron Ore — Brazil 7,025 2,622 9,646 7.55 4.76 227 42.5

Manganese 3,687 - 3,687 2.88 1.82 718 5.1


Thermal Coal 6,828 359 7,187 5.62 3.55 1,458 4.9



Less Minorities (25,902) (20.27) (12.79)



Equity Value 73,350 57.39 36.22



Exchange Rate 1.58




AND LOVE THE ORE

Glencore Xstrata Plc Pro Forma The combined Glencore Xstrata entity (pro forma analysis) shows the greatest downside and least upside in the Grizzly and De Niro scenarios. This is because Glencore Xstrata is the play for investors who want sector exposure with minimal iron ore (Xstrata has none, Glencore has in 1Q:13 acquired a stake in Carl Ichan-backed Ferrous Resources, which produced just 3.2Mt in 2012 and plans to boost production to 5Mt in 2013, and reach 17Mt by 2016). Note that while we have outperform ratings on both Glencore and Xstrata, we have no rating on the combined entity as it remains pro forma until its listing.

Exhibit 378 Glencore Xstrata Pro Forma Scenario Analysis Summary

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 379 Glencore Xstrata Pro Forma EV by Division


Glencore Xstrata Pro Forma NPV per share vs. Current Model


Grizzly 1.21 -77%

Bear 3.05 -41%

Bull 4.90 -5%

De Niro 6.75 31%

150,387

62,457

103,780

145,426

187,125

(20,000)

-

20,000

40,000

60,000

80,000

100,000

120,000

140,000

160,000

180,000

200,000


US

$m

Glencore Xstrata Pro-forma EV

Zinc Copper NickelAlumina & Aluminium Coal Oil & Oil ProductsAgricultural Products Alloys Marketing - Metals & MiningMarketing - Energy Products Marketing - Agricultural Products Synergies




Exhibit 380 Glencore Xstrata Pro Forma — Current Model — Pro Forma Financials



December year-end 2008 E 2009 E 2010 E 2011 E 2012 E 2013 E 2014 E 2015 E

Income Statement

Group revenue US$m 173,369 123,847 169,189 211,548 229,505 251,245 283,428 299,966

Operating profit US$m 11,302 6,546 10,696 11,325 8,394 14,944 21,432 23,908

PAT US$m 7,841 4,310 7,202 8,222 6,164 11,023 16,314 18,689

Minority interests US$m (339) (296) (475) (456) (497) (1,107) (1,497) (1,572)

Reported Earnings US$m 7,502 4,014 6,727 7,766 5,666 9,915 14,816 17,117

Reported EPS, USD/sh US$/share 0.56 0.30 0.50 0.58 0.43 0.74 1.11 1.28

Balance Sheet

Total assets US$m 110,935 122,488 139,637 149,128 157,463 167,726 176,720 186,355

Total liabilities US$m (75,915) (77,237) (84,948) (82,961) (83,581) (84,645) (81,292) (76,704)

Net assets US$m 35,020 45,251 54,689 66,167 73,882 83,081 95,428 109,651

Total equity US$m 35,020 45,251 54,689 66,167 73,882 83,081 95,428 109,651

Net debt/(Net cash) US$m 33,979 34,813 36,895 35,384 34,881 28,181 15,927 (18)

Cash Flow Statement

EBITDA 0 14,638 9,949 14,708 15,878 13,784 22,217 29,264 31,608

Operating cash flow US$m 17,322 3,870 11,935 13,012 15,464 18,028 27,492 30,724

Net cash flow from operations US$m 14,640 2,312 9,262 10,724 12,678 14,650 22,355 24,631

Net cash flow from investing US$m (13,128) (6,842) (7,913) (12,244) (8,872) (6,182) (4,835) (3,262)

Net cash flow from financing US$m (929) 4,631 (282) 2,321 (5,157) (6,184) (12,485) (13,664)

Closing cash & equivalents 31 Dec US$m 1,982 2,037 3,185 3,253 2,130 4,414 9,448 17,154

FCF Per Share US$/share 0.64 -0.14 0.19 0.05 0.18 0.62 1.26 1.51



AND LOVE THE ORE

Exhibit 381 Glencore Xstrata Pro Forma — Current Model — Valuation



Zinc 16,070 1,807 17,877 1.34 .84 3,097 5.8

Copper 54,408 6,491 60,899 4.57 2.85 7,989 7.6

Nickel 8,233 1,375 9,608 .72 .45 1,196 8.0

Alumina & Aluminium 377 - 377 .03 .02 51 7.4

Coal 25,583 3,535 29,118 2.19 1.36 5,489 5.3

Oil & Oil Products 3,249 262 3,511 .26 .16 572 6.1

Agricultural Products 772 - 772 .06 .04 283 2.7

Alloys 390 - 390 .03 .02 275 1.4

Marketing - Metals & Mining 8,340 1,157 9,497 .71 .44 1,309 7.3

Marketing - Energy Products 5,364 873 6,237 .47 .29 791 7.9

Marketing - Agricultural Products 5,345 801 6,146 .46 .29 919 6.7

Synergies 5,097 860 5,956 .45 .28 750 7.9

With Synergies






Equity Value 110,343 8.28 5.16



Exchange Rate 1.60


Without Synergies






Equity Value 104,387 7.83 4.88



Exchange Rate 1.60





Exhibit 382 Glencore Xstrata Pro Forma — Grizzly — Pro Forma Financials


Income Statement


Income Statement



PAT US$m 7,864 4,333 7,225 8,245 6,219 1,491 1,053 1,723

Minority interests US$m (339) (296) (475) (456) (497) (367) (377) (386)

Reported earnings US$m 7,525 4,037 6,750 7,789 5,722 1,124 676 1,337

Reported EPS US$/share 0.56 0.30 0.51 0.58 0.43 0.08 0.05 0.10

Balance Sheet

Total assets US$m 110,507 122,060 139,209 148,700 156,874 154,053 154,035 154,057


Net assets US$m 34,592 44,823 54,261 65,739 73,473 73,767 74,057 75,254

Total equity US$m 34,592 44,823 54,261 65,739 73,473 73,767 74,057 75,254

Net debt/(Net cash) US$m 33,979 34,813 36,895 35,384 35,482 33,314 31,446 27,947

Cash Flow Statement

EBITDA 0 14,638 9,949 14,708 15,878 13,858 10,271 10,320 11,138






FCF per Share US$/share 0.64 -0.14 0.19 0.05 0.19 0.36 0.25 0.38



AND LOVE THE ORE

Exhibit 383 Glencore Xstrata — Grizzly — Valuation



Zinc 6,533 1,433 7,966 .60 .37 2,000 4.0

Copper 21,775 4,526 26,301 1.97 1.23 4,595 5.7

Nickel 5,000 1,318 6,318 .47 .30 663 9.5

Alumina & Aluminum (413) - (413) (.03) (.02) 13 (32.8)

Coal 84 1,363 1,448 .11 .07 3,173 .5

Oil & Oil Products 694 113 808 .06 .04 419 1.9

Agricultural Products (3,579) - (3,579) (.27) (.17) 142 (25.2)

Alloys (1,115) - (1,115) (.08) (.05) 155 (7.2)

Marketing — Metals & Mining 6,731 1,065 7,795 .58 .36 1,289 6.0

Marketing — Energy Products 4,375 739 5,114 .38 .24 717 7.1

Marketing — Agricultural Products 4,973 759 5,732 .43 .27 220 26.0

Synergies 5,174 908 6,082 .46 .28 750 8.1

With Synergies






Equity Value 25,761 1.93 1.21



Exchange Rate 1.60


Without Synergies






Equity Value 19,679 1.48 .93



Exchange Rate 1.58

Per Share NPV (GBP) .93




Exhibit 384 Glencore Xstrata Pro Forma — Bear — Pro Forma Financials


Income Statement


Income Statement



PAT US$m 7,864 4,333 7,225 8,245 6,219 5,690 5,731 6,908

Minority interests US$m (339) (296) (475) (456) (497) (711) (744) (769)

Reported earnings US$m 7,525 4,037 6,750 7,789 5,722 4,979 4,987 6,138


Balance Sheet

Total assets US$m 110,507 122,060 139,209 148,700 156,874 159,665 161,230 163,188


Net assets US$m 34,592 44,823 54,261 65,739 73,473 77,428 81,040 86,042

Total equity US$m 34,592 44,823 54,261 65,739 73,473 77,428 81,040 86,042


Cash Flow Statement

EBITDA 0 14,638 9,949 14,708 15,878 13,858 15,466 16,041 17,336









AND LOVE THE ORE

Exhibit 385 Glencore Xstrata Pro Forma — Bear — Valuation



Zinc 11,248 1,983 13,231 .99 .62 2,000 6.6

Copper 33,347 6,138 39,485 2.96 1.85 4,595 8.6

Nickel 7,384 1,748 9,133 .69 .43 663 13.8

Alumina & Aluminum (147) - (147) (.01) (.01) 13 (11.7)

Coal 10,037 3,222 13,259 .99 .62 3,173 4.2

Oil & Oil Products 1,819 184 2,003 .15 .09 419 4.8

Agricultural Products 103 - 103 .01 .00 142 .7

Alloys 178 301 479 .04 .02 155 3.1




Synergies 5,077 865 5,942 .45 .28 750 7.9

With Synergies






Equity Value 65,142 4.89 3.05



Exchange Rate 1.60


Without Synergies






Equity Value 59,200 4.44 2.80



Exchange Rate 1.58





Exhibit 386 Glencore Xstrata — Bull — Pro Forma Financials


Income Statement


Income Statement



PAT US$m 7,864 4,333 7,225 8,245 6,219 9,914 10,453 12,140




Balance Sheet

Total assets US$m 110,507 122,060 139,209 148,700 156,874 165,392 168,691 172,699


Net assets US$m 34,592 44,823 54,261 65,739 73,473 81,302 88,517 97,537

Total equity US$m 34,592 44,823 54,261 65,739 73,473 81,302 88,517 97,537


Cash Flow Statement

EBITDA 0 14,638 9,949 14,708 15,878 13,858 20,665 21,767 23,540









AND LOVE THE ORE

Exhibit 387 Glencore Xstrata Pro Forma — Bull — Valuation



Zinc 15,958 2,568 18,526 1.39 .87 2,000 9.3

Copper 44,942 7,852 52,794 3.96 2.47 4,595 11.5

Nickel 9,780 2,210 11,990 .90 .56 663 18.1


Coal 19,905 5,088 24,993 1.88 1.17 3,173 7.9

Oil & Oil Products 2,942 256 3,198 .24 .15 419 7.6

Agricultural Products 3,730 - 3,730 .28 .17 142 26.2

Alloys 1,499 626 2,125 .16 .10 155 13.7




Synergies 5,045 846 5,891 .44 .28 750 7.9

With Synergies






Equity Value 104,634 7.85 4.90



Exchange Rate 1.60


Without Synergies






Equity Value 98,743 7.41 4.68



Exchange Rate 1.58





Exhibit 388 Glencore Xstrata — De Niro — Pro Forma Financials


Income Statement


Income Statement



PAT US$m 7,864 4,333 7,225 8,245 6,219 14,143 15,181 17,379




Balance Sheet

Total assets US$m 110,507 122,060 139,209 148,700 156,874 171,122 176,157 182,467


Net assets US$m 34,592 44,823 54,261 65,739 73,473 85,184 96,005 109,047

Total equity US$m 34,592 44,823 54,261 65,739 73,473 85,184 96,005 109,047

Net debt/(Net cash) US$m 33,979 34,813 36,895 35,384 34,261 27,719 15,467 (82)

Cash Flow Statement

EBITDA 0 14,638 9,949 14,708 15,878 13,858 25,866 27,496 29,748









AND LOVE THE ORE

Exhibit 389 Glencore Xstrata — De Niro — Valuation



Zinc 20,660 3,164 23,824 1.79 1.11 2,000 11.9

Copper 56,528 9,601 66,129 4.96 3.09 4,595 14.4

Nickel 12,175 2,683 14,857 1.11 .70 663 22.4


Coal 29,730 6,944 36,674 2.75 1.72 3,173 11.6

Oil & Oil Products 4,063 330 4,394 .33 .21 419 10.5

Agricultural Products 7,335 - 7,335 .55 .34 142 51.6

Alloys 2,848 953 3,801 .29 .18 155 24.5


Marketing - Energy Products 5,727 1,011 6,738 .51 .32 717 9.4


Synergies 5,027 834 5,861 .44 .27 750 7.8

With Synergies


Less Minorities (10,038) (.75) (.47)




Equity Value 144,197 10.82 6.75



Exchange Rate 1.60


Without Synergies


Less Minorities (10,038) (.75) (.48)




Equity Value 138,337 10.38 6.55



Exchange Rate 1.58





Vale Vale is the "pure" iron ore play in our coverage, with 70% of FY 2012 revenues coming from iron ore. As such, and given iron ore’s historical volatility, Vale has a downside tied with Rio Tinto in the Grizzly and Bear Scenarios, but the greatest upside in the Bull and De Niro scenarios. Vale remains our least preferred stock — the company incurs more capex/ton for its new iron ore exposure than Rio Tinto, and we question how the significant involvement of the Brazilian government (directly, through Golden Shares, and indirectly, through Valepar) influences the company’s strategic direction vis-a-vis shareholder interests.

Exhibit 390 Vale Scenario Analysis Summary

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 391 Vale Pro Forma EV by Division


Vale Implied Share Price (BRL) vs. Current Model


Grizzly 9.73 -67%

Bear 34.38 18%

Bull 58.97 102%

De Niro 83.54 186%

102,468

51,601

115,951

180,127

244,263

(50,000)

-

50,000

100,000

150,000

200,000

250,000

300,000


US

$m

Vale EV

Iron Ore Manganese&Ferroalloys Coal Base Metals Fertilizers Other



AND LOVE THE ORE

Exhibit 392 Vale — Current Model — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 393 Vale — Current Model — Valuation


Pro-Forma Financials


Income Statement

Net Operating Revenue US$m 19,647 32,242 37,426 23,311 45,293 60,947 47,694 59,716 74,202 63,530


PAT US$m 6,782 12,166 15,123 4,546 17,493 21,140 11,281 17,802 26,784 17,080

Minority interests US$m (579) (802) (258) (107) (189) 233 257 630 1,006 726


Reported EPS - preferred, US$/sh US$/share 2.14 2.42 2.53 .97 3.23 4.37 1.07 3.73 5.48 3.51

Reported EPS - common, US$/sh US$/share 2.18 2.42 2.53 .97 3.22 4.34 1.07 3.73 5.48 3.51

Balance Sheet

Total assets US$m 60,926 76,717 79,992 102,279 129,139 128,728 131,478 145,353 166,232 178,561


Net assets US$m 22,484 35,831 45,047 60,497 72,441 80,114 76,363 90,871 112,097 125,616

Total equity US$m 22,484 35,831 45,047 60,497 72,441 80,114 76,363 90,871 112,097 125,616

Net debt/(Net Cash) US$m 18,108 17,978 8,216 16,320 18,253 20,860 26,088 26,361 19,152 10,244

Cash Flow Statement

EBITDA US$m 8,620 15,380 18,505 8,779 24,955 32,722 18,133 31,950 43,806 30,513




Net cash flow from financing US$m 13,345 (5,209) 9,004 625 (2,569) (14,371) 1,165 (3,716) (7,360) (5,788)


FCF Per Share US$/share 1.09 1.16 1.64 0.03 1.53 1.89 0.42 0.90 2.70 2.55

Valuation Summary

Price Target/History BRL 54.10 50.75 23.89 42.20 48.50 42.69 38.49 29.25 29.25 29.25


MCAP US$m 76,537 139,635 52,916 113,098 146,075 133,023 100,447 74,796 74,796 74,796

EV US$m 101,433 167,084 62,164 131,681 165,914 152,529 121,850 102,468 102,468 102,468

EV/EBITDA - 11.8 10.9 3.4 15.0 6.6 4.7 6.7 3.2 2.3 3.4

PE - 25.3 21.0 9.4 43.4 15.0 9.8 36.0 7.8 5.3 8.3


Iron Ore 69,020 15,363 84,383 16.55 32.99 27,326 3.1

Manganese & Ferroalloys 1,135 68 1,203 .24 .47 345 3.5

Coal 7,317 2,823 10,140 1.99 3.96 467 21.7

Base Metals 9,766 - 9,766 1.92 3.82 2,322 4.2

Fertilizers 5,078 - 5,078 1.00 1.99 1,917 2.6

Other (8,103) - (8,103) (1.59) (3.17) (428) 18.9





Equity Value 75,178 14.74 29.25



Exchange Rate 1.99





Exhibit 394 Vale — Grizzly — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 395 Vale — Grizzly — Valuation




Income Statement



PAT US$m 6,782 12,166 15,123 4,546 17,493 21,140 11,281 5,027 2,381 3,207

Minority interests US$m (579) (802) (258) (107) (189) 233 257 109 (104) (125)


Reported EPS - preferred, US$/sh US$/share 2.14 2.42 2.53 .97 3.23 4.37 1.07 1.11 .45 .61

Reported EPS - common, US$/sh US$/share 2.18 2.42 2.53 .97 3.22 4.34 1.07 1.11 .45 .61

Balance Sheet

Total assets US$m 60,926 76,717 79,992 102,279 129,139 128,728 131,478 136,284 141,169 145,257


Net assets US$m 22,484 35,831 45,047 60,497 72,441 80,114 76,363 80,756 82,681 85,272

Total equity US$m 22,484 35,831 45,047 60,497 72,441 80,114 76,363 80,756 82,681 85,272


Cash Flow Statement

EBITDA US$m 8,620 15,380 18,505 8,779 24,955 32,722 18,133 14,342 10,753 12,379




Net cash flow from financing US$m 13,345 (5,209) 9,004 625 (2,569) (14,371) 1,165 (22) 963 (328)

Closing cash & equivalents US$m 4,448 1,046 10,331 7,293 7,584 3,531 5,832 2,522 (1,734) (2,599)

FCF Per Share US$/share 1.09 1.16 1.64 0.03 1.53 1.89 0.42 -0.44 -0.76 0.20

Valuation Summary



MCAP US$m 76,537 139,635 52,916 113,098 146,075 133,023 100,447 24,759 24,759 24,759

EV US$m 101,433 167,084 62,164 131,681 165,914 152,529 121,850 51,601 51,601 51,601

EV/EBITDA - 11.8 10.9 3.4 15.0 6.6 4.7 6.7 3.6 4.8 4.2

PE - 25.3 21.0 9.4 43.4 15.0 9.8 36.0 8.8 21.7 16.0


Iron Ore 40,155 13,197 53,352 10.46 20.86 17,647 3.0

Manganese & Ferroalloys 506 41 547 .11 .21 190 2.9

Coal (293) 1,503 1,210 .24 .47 (449) (2.7)

Base Metals 763 151 914 .18 .36 603 1.5

Fertilizers 2,608 - 2,608 .51 1.02 632 4.1

Other (7,030) - (7,030) (1.38) (2.75) (490) 14.3


Less Minorities (627) (.12) (.25)



Equity Value 24,886 4.88 9.73



Exchange Rate 1.99




AND LOVE THE ORE

Exhibit 396 Vale — Bear — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 397 Vale — Bear — Valuation




Income Statement



PAT US$m 6,782 12,166 15,123 4,546 17,493 21,140 11,281 9,327 8,774 10,416

Minority interests US$m (579) (802) (258) (107) (189) 233 257 430 337 323




Balance Sheet

Total assets US$m 60,926 76,717 79,992 102,279 129,139 128,728 131,478 139,311 148,331 157,044


Net assets US$m 22,484 35,831 45,047 60,497 72,441 80,114 76,363 84,131 91,083 99,351

Total equity US$m 22,484 35,831 45,047 60,497 72,441 80,114 76,363 84,131 91,083 99,351


Cash Flow Statement

EBITDA US$m 8,620 15,380 18,505 8,779 24,955 32,722 18,133 20,399 19,668 22,302







Valuation Summary



MCAP US$m 76,537 139,635 52,916 113,098 146,075 133,023 100,447 87,494 87,494 87,494

EV US$m 101,433 167,084 62,164 131,681 165,914 152,529 121,850 115,951 115,951 115,951

EV/EBITDA - 11.8 10.9 3.4 15.0 6.6 4.7 6.7 5.7 5.9 5.2

PE - 25.3 21.0 9.4 43.4 15.0 9.8 36.0 17.0 19.1 16.2


Iron Ore 81,692 23,544 105,236 20.64 41.14 17,647 6.0

Manganese & Ferroalloys 886 89 974 .19 .38 190 5.1

Coal 3,112 2,575 5,687 1.12 2.22 (449) (12.7)

Base Metals 6,349 1,115 7,464 1.46 2.92 603 12.4

Fertilizers 5,057 - 5,057 .99 1.98 632 8.0

Other (8,468) - (8,468) (1.66) (3.31) (490) 17.3





Equity Value 87,942 17.24 34.38



Exchange Rate 1.99





Exhibit 398 Vale — Bull — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 399 Vale — Bull — Valuation




Income Statement



PAT US$m 6,782 12,166 15,123 4,546 17,493 21,140 11,281 13,712 15,323 17,965

Minority interests US$m (579) (802) (258) (107) (189) 233 257 752 777 771




Balance Sheet

Total assets US$m 60,926 76,717 79,992 102,279 129,139 128,728 131,478 142,390 155,636 169,177


Net assets US$m 22,484 35,831 45,047 60,497 72,441 80,114 76,363 87,575 99,678 113,895

Total equity US$m 22,484 35,831 45,047 60,497 72,441 80,114 76,363 87,575 99,678 113,895


Cash Flow Statement

EBITDA US$m 8,620 15,380 18,505 8,779 24,955 32,722 18,133 26,456 28,583 32,224







Valuation Summary



MCAP US$m 76,537 139,635 52,916 113,098 146,075 133,023 100,447 150,060 150,060 150,060

EV US$m 101,433 167,084 62,164 131,681 165,914 152,529 121,850 180,127 180,127 180,127

EV/EBITDA - 11.8 10.9 3.4 15.0 6.6 4.7 6.7 6.8 6.3 5.6

PE - 25.3 21.0 9.4 43.4 15.0 9.8 36.0 20.0 18.6 16.0


Iron Ore 123,153 33,829 156,982 30.78 61.37 17,647 8.9


Coal 6,510 3,640 10,151 1.99 3.97 (449) (22.6)

Base Metals 11,926 2,071 13,997 2.74 5.47 603 23.2

Fertilizers 7,501 - 7,501 1.47 2.93 632 11.9

Other (9,905) - (9,905) (1.94) (3.87) (490) 20.2


Less Minorities (3,212) (.63) (1.26)



Equity Value 150,827 29.57 58.97



Exchange Rate 1.99




AND LOVE THE ORE

Exhibit 400 Vale — De Niro — Pro Forma Financials

Source: Corporate reports and Bernstein estimates and analysis. Exhibit 401 Vale — De Niro — Valuation




Income Statement



PAT US$m 6,782 12,166 15,123 4,546 17,493 21,140 11,281 18,223 22,081 25,655

Minority interests US$m (579) (802) (258) (107) (189) 233 257 1,073 1,218 1,219




Balance Sheet

Total assets US$m 60,926 76,717 79,992 102,279 129,139 128,728 131,478 145,544 163,142 181,594


Net assets US$m 22,484 35,831 45,047 60,497 72,441 80,114 76,363 91,120 108,541 128,821

Total equity US$m 22,484 35,831 45,047 60,497 72,441 80,114 76,363 91,120 108,541 128,821


Cash Flow Statement

EBITDA US$m 8,620 15,380 18,505 8,779 24,955 32,722 18,133 32,513 37,499 42,146







Valuation Summary



MCAP US$m 76,537 139,635 52,916 113,098 146,075 133,023 100,447 212,585 212,585 212,585

EV US$m 101,433 167,084 62,164 131,681 165,914 152,529 121,850 244,263 244,263 244,263

EV/EBITDA - 11.8 10.9 3.4 15.0 6.6 4.7 6.7 7.5 6.5 5.8

PE - 25.3 21.0 9.4 43.4 15.0 9.8 36.0 21.4 18.2 15.8


Iron Ore 164,597 44,101 208,698 40.92 81.59 17,647 11.8


Coal 9,908 4,704 14,611 2.87 5.71 (449) (32.5)

Base Metals 17,498 3,025 20,523 4.02 8.02 603 34.0

Fertilizers 9,944 - 9,944 1.95 3.89 632 15.7

Other (11,343) - (11,343) (2.22) (4.43) (490) 23.1


Less Minorities (4,503) (.88) (1.76)



Equity Value 213,672 41.90 83.54



Exchange Rate 1.99





Complete Commodity Price Scenarios

Exhibit 402 Long-Term Real Commodity Price Scenarios


Real LT Commodity prices Scenario as % of Current LT Price

Current Model


Platinum US$/oz 1,486 1,500 1,700 1,900 2,100 101% 114% 128% 141%

Palladium US$/oz 606 500 600 700 800 83% 99% 116% 132%

Rhodium US$/oz 1,110 1,000 1,200 1,400 1,600 90% 108% 126% 144%

Gold US$/oz 1,619 1,200 1,500 1,800 2,100 74% 93% 111% 130%

Copper US$/t 7,652 6,500 7,500 8,500 9,500 85% 98% 111% 124%

Nickel US$/t 15,991 16,000 18,000 20,000 22,000 100% 113% 125% 138%

Zinc US$/t 1,841 1,800 2,000 2,200 2,400 98% 109% 119% 130%

Lead US$/t 2,003 1,800 2,000 2,200 2,400 90% 100% 110% 120%

Hard Coking Coal US$/t 194 150 180 210 240 77% 93% 108% 124%

PCI US$/t 158 115 138 160 183 73% 87% 102% 116%

Semi-Soft Coking Coal US$/t 149 107 128 149 171 72% 86% 100% 115%

Thermal Coal ex Newcastle US$/t 88 80 90 100 110 91% 103% 114% 126%

Thermal Coal ex RBCT US$/t 89 81 92 102 112 91% 103% 114% 126%

Thermal Coal ex Colombia US$/t 82 75 85 94 103 91% 103% 114% 126%

Iron Ore Fines FOB Aus US$/t 74 70 90 110 130 94% 121% 148% 175%

Iron Ore Lump FOB Aus. US$/t 79 74 96 117 138 94% 121% 148% 175%

Iron Ore Fines FOB Brazil US$/t 69 65 84 103 121 94% 121% 148% 175%

Iron Ore Concentrate FOB Can US$/t 64 60 77 94 111 94% 121% 148% 175%

Iron Ore Pellets FOB Canada US$/t 81 76 98 120 142 94% 121% 148% 175%

Diamonds US$/carat 77 80 85 90 95 104% 111% 117% 124%

Phosphates US$/t 620 600 675 750 825 97% 109% 121% 133%

Niobium US$/t 36,485 35,000 40,000 45,000 50,000 96% 110% 123% 137%

Manganese Ore US$/t 222 200 230 260 290 90% 104% 117% 131%

Ferro Manganese US$/t 1,148 1,034 1,189 1,344 1,499 90% 104% 117% 131%

Ferro Chrome USc/lb 101 100 110 120 130 99% 109% 119% 129%

Ferro Vanadium US$/kg 29 25 30 35 40 87% 105% 122% 140%

V2O5 US$/lb 5 5 6 7 8 86% 105% 124% 143%

Silver US$/oz 30 20 25 30 35 67% 83% 100% 117%

Cobalt US$/lb 13 10 15 20 25 80% 120% 160% 200%

Uranium US$/lb 44 45 50 55 60 101% 113% 124% 135%

Molybdenum US$/lb 11 10 12 14 16 91% 109% 127% 145%

Bauxite US$/t 64 40 50 60 70 62% 78% 94% 109%

Alumina US$/t 373 273 305 337 369 73% 82% 90% 99%

Aluminum US$/t 2,405 1,700 1,900 2,100 2,300 71% 79% 87% 96%

Oil WTI US$/bbl 122 85 95 105 115 70% 78% 86% 94%

Natural Gas - Henry Hub US$/mmbtu 5 3 4 5 6 48% 67% 87% 106%

Tin US$/t 19,601 17,381 20,055 22,729 25,403 89% 102% 116% 130%

Brent US$/bbl 122 98 110 122 133 81% 90% 99% 109%

Nat Gas - Japan US$/mmbtu 13 11 16 21 25 90% 127% 163% 199%

Blended Soy Meal & Oil US$/t 776 600 750 900 1,050 77% 97% 116% 135%

German rapeseed US/t 748 578 723 867 1,012 77% 97% 116% 135%

Ukrainian oilseed US/t 554 428 535 642 750 77% 97% 116% 135%

Sugar Spot US/t 563 435 544 653 762 77% 97% 116% 135%

Flour (Baltic) US/t 40 31 39 47 55 77% 97% 116% 135%

Cotton (Brazilian) US/t 2,068 1,599 1,998 2,398 2,798 77% 97% 116% 135%

Corn (US Export) US/t 242 187 234 281 327 77% 97% 116% 135%



AND LOVE THE ORE

History of Commodity Prices

Exhibit 403 Nickel Spot Price Distribution

Source: Bloomberg L.P. and Bernstein analysis. Exhibit 404 Zinc Spot Price Distribution


0%

2%

4%

6%

8%

10%

12%

14%

Pe

rce

ntil

e

Ni Spot Price (US$/t)

Five-Year Nominal Nickel Price Distribution

Spot as of 04/22/2013 $15,099/t

0%

5%

10%

15%

20%

25%

1,0001,1001,2001,3001,4001,5001,6001,7001,8001,9002,0002,1002,2002,3002,4002,5002,6002,7002,800

Pe

rce

ntil

e

Zn Spot Price (US$/t)

Five-Year Nominal Zinc Price Distribution

Spot as of 04/22/2013 $1,851/t




Exhibit 405 Aluminum Spot Price Distribution

Source: Bloomberg L.P. and Bernstein analysis. Exhibit 406 Platinum Spot Price Distribution


0%

2%

4%

6%

8%

10%

12%

14%

16%

Pe

rce

ntil

e

Al Spot Price (US$/t)

Five-Year Nominal Aluminum Price Distribution

Spot as of 04/22/2013 $1,872/t

0%

2%

4%

6%

8%

10%

12%

14%

750

800

850

900

950

1,00

0

1,05

0

1,10

0

1,15

0

1,20

0

1,25

0

1,30

0

1,35

0

1,40

0

1,45

0

1,50

0

1,55

0

1,60

0

1,65

0

1,70

0

1,75

0

1,80

0

1,85

0

1,90

0

1,95

0

2,00

0

2,05

0

2,10

0

2,15

0

2,20

0

2,25

0

2,30

0

2,35

0

Pe

rce

ntil

e

Pt Spot Price (US$/oz)

Five-Year Nominal Platinum Price Distribution

Spot as of 04/22/2013 $1,439/oz



AND LOVE THE ORE

Exhibit 407 Palladium Spot Price Distribution

Source: Bloomberg L.P. and Bernstein analysis. Exhibit 408 Gold Spot Price Distribution


0%

2%

4%

6%

8%

10%

12%

14%

150 200 250 300 350 400 450 500 550 600 650 700 750 800 850

Pe

rce

ntil

e

Pl Spot Price (US$/oz)

Five-Year Nominal Palladium Price DistributionSpot as of 04/22/2013 $679/oz

0%

1%

2%

3%

4%

5%

6%

7%

8%

Pe

rce

ntil

e

Au Spot Price (US$/oz)

Five-Year Nominal Gold Price Distribution

Spot as of 04/22/2013 $1,414/oz




Exhibit 409 Silver Spot Price Distribution

Source: Bloomberg L.P. and Bernstein analysis. Exhibit 410 Thermal Coal Spot Price Distribution


0%

5%

10%

15%

20%

25%

30%

35%

5 10 15 20 25 30 35 40 45 50

Pe

rce

ntil

e

Ag Spot Price (US$/oz)

Five-Year Nominal Silver Price Distribution

Spot as of 04/22/2013 $23.66/oz

0%

2%

4%

6%

8%

10%

12%

14%

60 65 70 75 80 85 90 95 100105110115120125130135140145150155160165170175180185190195200

Pe

rce

ntil

e

Spot Price (US$/t)

Five-Year Nominal Thermal Coal FOB Aus Price Distribution

Spot as of 04/22/2013

$86/t



AND LOVE THE ORE

Exhibit 411 Molybdenum Spot Price Distribution

Source: Bloomberg L.P. and Bernstein analysis. Exhibit 412 Uranium Spot Price Distribution


0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

20%

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

Pe

rce

ntil

e

Spot Price (US$/lb)

Five-Year Nominal Molybdenum Price Distribution

Spot as of 04/22/2013 $11.3/lb

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 92

Pe

rce

ntil

e

Spot Price (US$/lb)

Five-Year Nominal Uranium Price Distribution

Spot as of 04/22/2013 $40.9/lb




Appendix 2: Growth Scenarios

This appendix provides full details on the internally consistent scenarios in the "What GDP Growth Is Consistent With $80/t Iron Ore Baked Into Mining Equities?" and "China Growth Scenarios" chapters. It shows the relationship between Chinese GDP, supply, demand and resultant iron ore price in five scenarios. These scenarios are:

1) Consensus implied scenario ($80/t iron ore) 2) April 2013 IMF GDP forecast 3) "Back to the Future" based upon IMF 2012 forecast 4) "Out With a Whimper" taking the Chinese official GDP target 5) "It's All Over Now Baby Blue," with 6% GDP growth and no supply

curtailment


254 E

UR

OP

EA

N M

ET

AL

S &

MIN

ING

: A S

TR

AN

GE

LO

VE

— H

OW

I LE

AR

NE

D T

O S

TO

P W

OR

RY

ING

AN

D L

OV

E T

HE

OR

E

Exhibit 413 Consensus Scenario


Chinese Crude Steel Production Summary 2005A 2006A 2007A 2008A 2009A 2010A 2011A 2012A 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E


China GDP Growth % 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 7.8% 7.6% 7.5% 7.2% 7.1% 7.0% 7.0% 7.0%





China Finished Steel Demand Growth % 25.2% 8.5% 11.8% 2.5% 26.0% 4.9% 9.0% 5.0% 4.0% 4.9% 3.7% 3.0% 2.7% 2.7% 2.7% 2.7%

Chinese Self Sufficiency in Steel % 98.9% 107.6% 110.7% 108.8% 99.8% 103.9% 103.9% 103.9% 104.2% 104.6% 104.9% 105.2% 105.5% 105.7% 106.0% 106.4%

Chinese Finished Steel Production Mt 348 410 472 475 550 600 655 687 717 755 786 812 836 860 886 913


China Crude Steel Production Mt 353 419 489 500 574 627 683 717 751 794 828 858 885 914 943 973

China Crude Steel Production Growth % 24.9% 18.7% 16.7% 2.3% 14.6% 9.3% 9.0% 5.0% 4.3% 5.4% 4.0% 3.3% 2.9% 3.0% 3.0% 3.0%





Chinese Steel Cpaital Stock/Capita kg/Capita 2,023 2,265 2,534 2,803 3,150 3,508 3,896 4,295 4,703 5,124 5,552 5,994 6,439 6,889 7,343 7,801

Steel Growth to GDP Growth Ratio - 2.2 0.7 0.8 0.3 2.8 0.5 1.0 0.6 0.5 0.6 0.5 0.4 0.4 0.4 0.4 0.4


Assumed Consensus Steel Production Mt 353 419 489 500 574 627 683 717 751 794 828 858 885 914 943 973



Assumed Consensus Iron Ore Demand Mt 538 648 751 739 870 930 999 1,034 1,061 1,098 1,133 1,154 1,175 1,200 1,232 1,269

Implied Iron Ore Price - Nominal US$/t 33 42 46 76 81 147 168 128 127 117 110 106 100 95 90 90

Implied Iron Ore Price - Real US$/t 38 48 52 84 88 156 175 131 127 115 106 100 92 86 80 78


SO

N LA

PO

RT

E at P

ER

RY

CA

PIT

AL on 16-Jul-2013

E

UR

OP

EA

N M

ET

AL

S &

MIN

ING

: A S

TR

AN

GE

LO

VE

— H

OW

I LE

AR

NE

D T

O S

TO

P W

OR

RY

ING

AN

D L

OV

E T

HE

OR

E 255

Exhibit 414 IMF Scenario




China GDP Growth % 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 8.0% 8.2% 8.5% 8.5% 8.5% 8.5% 8.1% 7.8%





China Finished Steel Demand Growth % 25.2% 8.5% 11.8% 2.5% 26.0% 4.9% 9.0% 5.0% 7.8% 9.3% 9.2% 5.8% 4.0% 2.9% -1.0% -1.6%


Chinese Finished Steel Production Mt 348 410 472 475 550 600 655 687 743 816 893 947 988 1,019 1,011 998


China Crude Steel Production Mt 353 419 489 500 574 627 683 717 779 857 941 1,001 1,046 1,082 1,076 1,064

China Crude Steel Production Growth % 24.9% 18.7% 16.7% 2.3% 14.6% 9.3% 9.0% 5.0% 8.1% 9.8% 9.5% 6.1% 4.2% 3.2% -0.8% -1.3%






Steel Growth to GDP Growth Ratio - 2.2 0.7 0.8 0.3 2.8 0.5 1.0 0.6 1.0 1.1 1.1 0.7 0.5 0.3 -0.1 -0.2









SO

N LA

PO

RT

E at P

ER

RY

CA

PIT

AL on 16-Jul-2013

256 E

UR

OP

EA

N M

ET

AL

S &

MIN

ING

: A S

TR

AN

GE

LO

VE

— H

OW

I LE

AR

NE

D T

O S

TO

P W

OR

RY

ING

AN

D L

OV

E T

HE

OR

E

Exhibit 415 Back to the Future




China GDP Growth % 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 8.2% 8.4% 8.7% 8.7% 8.7% 8.7% 8.3% 8.0%





China Finished Steel Demand Growth % 25.2% 8.5% 11.8% 2.5% 26.0% 4.9% 9.0% 5.0% 10.5% 9.3% 9.1% 5.7% 3.9% 2.9% -1.1% -1.7%


Chinese Finished Steel Production Mt 348 410 472 475 550 600 655 687 762 836 914 970 1,010 1,042 1,033 1,019


China Crude Steel Production Mt 353 419 489 500 574 627 683 717 798 878 964 1,025 1,070 1,106 1,099 1,086

China Crude Steel Production Growth % 24.9% 18.7% 16.7% 2.3% 14.6% 9.3% 9.0% 5.0% 10.8% 9.7% 9.4% 6.1% 4.2% 3.1% -0.8% -1.4%






Steel Growth to GDP Growth Ratio - 2.2 0.7 0.8 0.3 2.8 0.5 1.0 0.6 1.3 1.1 1.0 0.7 0.5 0.3 -0.1 -0.2









SO

N LA

PO

RT

E at P

ER

RY

CA

PIT

AL on 16-Jul-2013

E

UR

OP

EA

N M

ET

AL

S &

MIN

ING

: A S

TR

AN

GE

LO

VE

— H

OW

I LE

AR

NE

D T

O S

TO

P W

OR

RY

ING

AN

D L

OV

E T

HE

OR

E 257

Exhibit 416 Ending With a Whimper




China GDP Growth % 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 8.0% 8.0% 7.9% 7.7% 7.7% 7.7% 7.5% 7.5%





China Finished Steel Demand Growth % 25.2% 8.5% 11.8% 2.5% 26.0% 4.9% 9.0% 5.0% 7.3% 7.3% 4.2% 3.5% 4.0% 3.1% 0.9% 2.0%




China Crude Steel Production Mt 353 419 489 500 574 627 683 717 775 837 877 913 954 988 1,002 1,027

China Crude Steel Production Growth % 24.9% 18.7% 16.7% 2.3% 14.6% 9.3% 9.0% 5.0% 7.6% 7.7% 4.5% 3.8% 4.3% 3.3% 1.2% 2.3%






Steel Growth to GDP Growth Ratio - 2.2 0.7 0.8 0.3 2.8 0.5 1.0 0.6 0.9 0.9 0.5 0.5 0.5 0.4 0.1 0.3









SO

N LA

PO

RT

E at P

ER

RY

CA

PIT

AL on 16-Jul-2013

258 E

UR

OP

EA

N M

ET

AL

S &

MIN

ING

: A S

TR

AN

GE

LO

VE

— H

OW

I LE

AR

NE

D T

O S

TO

P W

OR

RY

ING

AN

D L

OV

E T

HE

OR

E

Exhibit 417 It's All Over Now Baby Blue




China GDP Growth % 11.3% 12.7% 14.2% 9.6% 9.2% 10.4% 9.3% 7.8% 7.6% 7.3% 7.0% 6.7% 6.3% 6.0% 6.0% 6.0%





China Finished Steel Demand Growth % 25.2% 8.5% 11.8% 2.5% 26.0% 4.9% 9.0% 5.0% 2.5% 2.1% 1.8% 1.6% -1.1% -0.4% 3.2% 2.7%




China Crude Steel Production Mt 353 419 489 500 574 627 683 717 740 761 780 797 792 792 822 848

China Crude Steel Production Growth % 24.9% 18.7% 16.7% 2.3% 14.6% 9.3% 9.0% 5.0% 2.8% 2.5% 2.1% 1.9% -0.8% -0.2% 3.5% 3.0%



Iron Ore Demand Mt 538 648 751 739 870 930 999 1,034 1,041 1,039 1,044 1,042 1,004 978 1,010 1,040



Steel Growth to GDP Growth Ratio - 2.2 0.7 0.8 0.3 2.8 0.5 1.0 0.6 0.3 0.3 0.3 0.2 -0.2 -0.1 0.5 0.5









SO

N LA

PO

RT

E at P

ER

RY

CA

PIT

AL on 16-Jul-2013



Appendix 3: Financial Statements

Exhibit 418 Rio Tinto Income Statement


Income Statement


Production

Iron Ore Mt 132.8 144.7 153.4 171.5 184.6 191.8 198.8 209.9 253.9 296.2

Copper kt 804 738 699 805 678 520 549 527 799 907

Gold koz 1,007 1,231 459 1,111 772 670 294 176 401 467

Coking Coal Mt 5.9 6.2 7.4 7.5 9.0 8.8 7.9 8.4 8.8 9.2

Thermal Coal Mt 29.1 22.2 19.3 20.2 18.4 17.4 19.7 24.2 24.2 22.9

Uranium klbs 12.6 12.6 14.2 14.1 11.4 7.1 9.8 9.5 9.5 10.6

Diamonds k carats 35,163 26,023 20,816 14,026 13,843 11,733 13,122 15,758 24,458 29,802

Aluminum kt 845 1,473 3,981 3,804 3,790 3,824 3,456 3,552 3,608 3,622

Prices

Iron Ore Fines US$/t 42 46 76 65 108 152 120 131 138 142

Copper US$/t 6,725 7,124 6,969 5,142 7,531 8,822 7,952 7,830 9,049 10,352

Gold US$/oz 604 695 872 972 1,225 1,571 1,669 1,541 1,572 1,604

Hard Coking Coal US$/t 117 103 328 145 220 296 210 179 203 222

Thermal Coal US$/t 49 66 129 72 99 121 94 92 104 116

Uranium US$/lb 47 99 64 47 46 56 49 43 49 54

Diamonds US$/carat 62 60 64 70 80 80 80 94 108 103

Aluminum US$/t 2,567 2,640 2,575 1,663 2,172 2,398 2,019 2,012 2,260 2,483

Revenue

Iron Ore US$m 6,938 9,193 16,150 12,598 24,024 29,909 24,695 27,958 34,937 41,194

Aluminum US$m 3,493 7,359 18,297 12,038 15,206 12,159 10,105 12,104 13,681 14,899

Copper US$m 7,079 8,501 5,829 6,206 7,782 7,555 6,520 5,788 9,786 12,481

Energy US$m 4,240 4,765 8,018 4,869 5,652 7,327 5,783 6,039 7,077 7,890

Diamonds & Minerals US$m 3,461 3,658 4,116 2,618 3,035 3,220 3,640 4,083 4,770 4,995

Other US$m 229 42 5,655 5,707 4,624 5,508 4,823 4,524 4,524 4,524

Group Revenue US$m 25,440 33,518 58,065 44,036 60,323 65,678 55,566 60,496 74,775 85,984

Associate Revenue US$m (2,975) (3,818) (3,801) (2,211) (3,747) (5,141) (4,599) (4,801) (5,534) (6,317)

Consolidated Revenue US$m 22,465 29,700 54,264 41,825 56,576 60,537 50,967 55,695 69,242 79,667

Operating Costs US$m (13,892) (20,752) (37,641) (33,818) (36,667) (36,260) (37,536) (38,769) (46,388) (50,135)

Exploration & Evaluation Costs US$m - (321) (1,134) (514) (594) (1,437) (1,970) (2,050) (2,533) (2,913)

Special Items and Other US$m 401 (56) (5,295) 13 379 (9,492) (15,071) - - -

Associate Income US$m 1,378 1,584 1,039 786 1,101 704 1,034 937 1,104 1,346

Operating Profit US$m 10,352 10,155 11,233 8,292 20,795 14,052 (2,576) 15,813 21,424 27,965

Net Interest US$m (54) (404) (1,414) (809) (615) (382) (160) (1,235) (1,181) (961)

Other Financing Charges US$m (58) 85 (641) 377 397 (456) 168 (390) (390) (390)

PBT US$m 10,240 9,836 9,178 7,860 20,577 13,214 (2,568) 14,188 19,853 26,614

Income Tax Expense US$m (2,373) (2,090) (3,742) (2,076) (5,296) (6,439) (429) (3,730) (5,413) (7,523)

Loss from Discontinued Operations US$m - - (827) (449) (97) (10) (7) - - -

Tax Rate % 23.2% 21.2% 40.8% 26.4% 25.7% 48.7% -16.7% 26.3% 27.3% 28.3%

PAT US$m 7,867 7,746 4,609 5,335 15,184 6,765 (3,004) 10,458 14,440 19,091

Minority Interests US$m (429) (434) (933) (463) (860) (939) 14 (614) (884) (1,506)

Reported Earnings US$m 7,438 7,312 3,676 4,872 14,324 5,826 (2,990) 9,844 13,555 17,586

Reported EPS, USD/sh USD/share 5.58 5.69 2.86 2.76 7.30 3.03 (1.62) 5.33 7.34 9.52

Underlying Earnings US$m 7,338 7,443 10,303 6,298 13,987 15,549 9,303 9,844 13,555 17,586

Underlying EPS, USD/sh USD/share 5.50 5.79 8.03 3.57 7.13 8.09 5.03 5.33 7.34 9.52

DPS USD/share 1.92 1.16 1.52 .68 .90 1.17 1.64 1.66 1.73 2.19

W. Average Shares Outstanding m 1,333 1,286 1,284 1,764 1,961 1,923 1,849 1,847 1,847 1,847



AND LOVE THE ORE

Exhibit 419 Rio Tinto Balance Sheet


Balance Sheet


Non-current assets

Intangible assets US$m 1,225 23,407 20,581 19,998 21,016 16,142 9,402 9,402 9,402 9,402

PP&E US$m 22,207 45,647 41,753 45,803 56,024 64,967 75,131 82,099 84,752 87,306

Investments in associates US$m 2,235 7,038 5,053 6,735 6,855 9,833 5,312 5,312 5,312 5,312

Other US$m 1,952 3,456 3,794 4,986 5,713 6,617 7,535 7,731 8,259 8,707

Total non-current assets US$m 27,619 79,548 71,181 77,522 89,608 97,559 97,380 104,544 107,725 110,727

Current assets

Inventories US$m 2,540 5,382 5,607 4,889 4,756 5,307 6,136 5,903 7,005 7,418

Receivables US$m 2,938 6,479 5,401 4,447 5,582 6,058 5,319 5,720 6,798 7,714

Cash and equivalents US$m 736 1,645 1,181 4,233 9,948 9,670 7,082 6,427 11,807 20,473

Other US$m 661 8,337 6,246 6,145 2,879 951 1,656 686 686 686

Total current assets US$m 6,875 21,843 18,435 19,714 23,165 21,986 20,193 18,736 26,296 36,291

Total assets US$m 34,494 101,391 89,616 97,236 112,773 119,545 117,573 123,280 134,022 147,018

Current liabilities

Payables US$m (2,693) (6,667) (7,197) (5,759) (6,570) (9,381) (9,244) (7,639) (9,065) (9,600)

Short-term debt US$m (1,504) (8,213) (10,034) (847) (2,151) (1,447) (2,228) (2,337) (1,440) (424)

Other US$m (1,583) (2,155) (2,748) (2,923) (4,155) (4,138) (2,349) (2,349) (2,349) (2,349)

Total current liabilities US$m (5,780) (17,035) (19,979) (9,529) (12,876) (14,966) (13,821) (12,325) (12,854) (12,373)

Non-current liabilities

Long-term debt US$m (2,007) (38,614) (29,724) (22,155) (13,277) (20,357) (24,591) (24,700) (23,803) (22,787)

Deferred tax liabilities US$m (2,425) (6,486) (4,054) (4,304) (5,222) (6,592) (5,119) (5,119) (5,119) (5,119)

Other US$m (4,897) (12,932) (13,398) (15,323) (16,886) (18,422) (16,021) (15,608) (15,608) (15,608)

Total non-current liabilities US$m (9,329) (58,032) (47,176) (41,782) (35,385) (45,371) (45,731) (45,427) (44,530) (43,514)


Net assets US$m 19,385 26,324 22,461 45,925 64,512 59,208 58,021 65,528 76,637 91,131

Equity

Shareholders equity US$m 18,232 24,772 20,638 43,831 58,247 52,539 46,865 54,372 65,481 79,975

Minority interests US$m 1,153 1,552 1,823 2,094 6,265 6,669 11,156 11,156 11,156 11,156

Total equity US$m 19,385 26,324 22,461 45,925 64,512 59,208 58,021 65,528 76,637 91,131





Exhibit 420 Rio Tinto Cash Flow Statement


Cash Flow


EBITDA 12,566 13,611 23,870 14,471 26,639 29,586 20,381 24,183 31,348 38,698

Change in working capital US$m (696) 129 431 991 (825) 347 401 (1,969) (1,282) (1,242)

Other US$m (674) (1,171) (3,633) (1,628) (2,284) (2,543) (4,234) (3,022) (3,608) (4,141)


Income tax paid US$m (2,799) (3,421) (3,899) (3,076) (4,100) (6,197) (5,823) (3,730) (5,413) (7,523)

Dividends paid to minorities US$m (193) (168) (348) (410) (457) (548) (422) (614) (884) (1,506)

Net interest paid US$m (128) (489) (1,538) (1,136) (696) (613) (837) (1,235) (1,181) (961)


Capital expenditure US$m (3,920) (5,000) (8,574) (5,388) (4,591) (12,335) (17,458) (12,707) (9,360) (9,535)

Acquisitions & investments US$m - (37,509) - (661) (1,061) (4,901) (1,647) - - -

Disposals US$m 14 32 2,734 2,677 4,027 491 944 557 - -

Other US$m (456) (265) (341) 15 (86) (93) (13) - - -


(Repayment)/receipt of debt US$m (619) 38,161 (7,970) (16,445) (9,360) 5,891 7,888 218 (1,793) (2,034)

Dividends paid to shareholders US$m (2,573) (1,507) (1,933) (876) (1,754) (2,236) (3,038) (2,337) (2,447) (3,091)

Shares issued US$m 31 37 23 14,877 342 424 2,945 - - -

Shares bought back US$m (2,370) (1,648) - - - (5,504) (1,471) - - -

Other financing activities US$m 142 54 772 (19) 162 (2,001) 1 - - -


Effects of exchange rate on cash US$m 30 (27) (101) (284) (139) (71) (49) - - -

Change in cash and equivalents US$m (1,645) 819 (507) 3,108 5,817 (303) (2,432) (655) 5,380 8,666

Closing cash and equivalents US$m 736 1,645 1,181 4,233 9,948 9,670 7,082 6,427 11,807 20,473

FCF to firm US$m 4,477 4,148 8,195 5,370 14,839 8,858 (6,733) 2,756 11,685 16,257

FCF per share US$/share 3.36 3.23 6.38 3.04 7.57 4.61 (3.64) 1.49 6.33 8.80



AND LOVE THE ORE

Exhibit 421 BHP Income Statement


Income Statement


Production

Crude oil, Condensate & NGL mmbbl 55 57 68 76 97 92 85 93 98 106

Natural Gas bcf 360 357 368 365 369 405 822 952 1,186 1,524

Alumina kt 4,187 4,460 4,554 4,396 3,841 4,010 4,152 5,114 5,735 5,838

Aluminum kt 1,362 1,340 1,298 1,233 1,241 1,246 1,153 1,209 1,284 1,284

Copper kt 1,268 1,250 1,376 1,207 1,075 1,139 1,095 1,175 1,258 1,315

Uranium t 3,936 3,486 4,144 4,007 2,279 4,045 3,885 4,112 4,112 4,112

Nickel kt 175 186 168 173 176 153 158 143 143 143

Iron Ore Mt 97 99 112 114 125 134 159 170 188 203

Manganese Ore kt 5,280 6,009 6,575 4,162 6,765 7,011 7,879 8,337 9,001 9,001

Manganese Alloy kt 652 662 775 467 602 774 688 490 413 413

Metallurgical Coal Mt 35 38 35 36 37 33 33 37 38 39

Energy Coal Mt 86 87 80 68 66 70 71 79 85 87

Prices

Oil - WTI US$/bbl 66 72 100 62 79 95 94 98 112 125

Natural Gas - Henry Hub US$/mmbtu 6.73 6.96 8.89 3.94 4.37 4.00 2.75 4.07 4.83 5.44

Aluminum US$/t 2,567 2,640 2,575 1,663 2,172 2,398 2,019 2,012 2,260 2,483

Copper US$/t 6,725 7,124 6,969 5,142 7,531 8,822 7,952 7,830 9,049 10,352

Uranium US$/lb 47 99 64 47 46 56 49 43 49 54

Nickel US$/t 24,102 37,147 21,155 14,655 21,799 22,896 17,544 16,507 17,890 19,412


Ferromanganese US$/t 853 1,379 2,957 1,316 1,395 1,323 1,223 1,128 1,264 1,398


Thermal Coal US$/t 49 66 129 72 99 121 94 92 104 116

Revenue

Petroleum US$m 5,230 5,885 8,382 7,211 8,782 10,737 12,937 14,198 18,935 24,417

Aluminum US$m 5,084 5,879 5,746 4,151 4,353 5,221 4,766 4,753 5,707 6,259

Base Metals US$m 10,294 12,635 14,774 7,105 10,409 14,152 11,596 12,317 13,958 16,351

Diamonds & Specialty Products US$m 1,263 893 969 896 1,272 1,517 1,326 681 499 499

Stainless Steel Materials US$m 2,955 6,901 5,088 2,355 3,617 3,861 2,993 2,454 2,504 2,714

Iron Ore US$m 4,782 5,524 9,455 10,048 11,139 20,412 22,601 20,125 24,177 27,314

Manganese US$m 1,037 1,244 2,912 2,536 2,150 2,423 2,152 2,050 2,311 2,498

Metallurgical Coal US$m 3,941 3,769 3,941 8,087 6,059 7,573 7,576 5,736 6,681 7,660

Energy Coal US$m 3,965 4,576 6,560 6,524 4,265 5,507 6,022 5,182 5,911 6,781

Other Group & Inter Segment US$m 548 167 1,646 1,298 752 336 257 100 100 100


Other income US$m 1,227 588 648 589 528 531 906 4,429 5,293 6,198

Operating costs US$m (25,655) (29,660) (35,976) (38,640) (33,296) (40,454) (49,216) (50,507) (55,003) (62,079)


Financial income US$m 226 260 293 309 215 245 225 128 100 253

Financial expense US$m (731) (650) (955) (852) (674) (806) (955) (1,508) (1,775) (1,644)

PBT US$m 14,166 18,011 23,483 11,617 19,571 31,255 23,186 20,138 29,397 37,318

Income tax expense US$m (3,207) (4,174) (6,798) (4,784) (6,112) (6,481) (7,238) (4,257) (7,088) (8,998)

Royalty-related taxation US$m (425) (341) (723) (495) (451) (828) (252) (1,393) (2,319) (2,944)

Tax rate % 25.6% 25.1% 32.0% 45.4% 33.5% 23.4% 32.3% 28.1% 32.0% 32.0%

PAT US$m 10,534 13,496 15,962 6,338 13,008 23,946 15,696 14,488 19,990 25,377




DPS USD/share .32 .39 .56 .82 .83 .91 1.10 1.14 .96 .99

Shares outstanding m 5,934 5,615 5,565 5,564 5,563 5,323 5,321 5,321 5,321 5,321




Exhibit 422 BHP Balance Sheet


Balance Sheet


Non-current assets

Intangible assets US$m 683 615 625 661 687 904 5,112 5,207 5,207 5,207

PP&E US$m 30,985 36,705 47,332 49,032 55,576 68,468 95,247 100,513 104,539 107,721

Other US$m 8,072 9,761 6,371 6,591 7,455 8,239 8,463 9,693 10,001 10,297


Current assets

Inventories US$m 2,732 3,296 4,971 4,821 5,334 6,154 6,233 6,850 7,544 8,360

Receivables US$m 3,831 4,689 9,801 5,153 6,543 8,197 7,704 8,299 9,888 11,412

Cash and equivalents US$m 776 1,937 4,237 10,833 12,456 10,084 4,781 3,748 10,028 20,943

Other US$m 1,437 1,165 2,671 1,679 801 845 1,733 1,982 1,982 1,982


Total assets US$m 48,516 58,168 76,008 78,770 88,852 102,891 129,273 136,292 149,188 165,922

Current liabilities

Payables US$m (4,053) (4,724) (6,774) (5,619) (6,467) (9,718) (12,024) (8,865) (9,763) (10,819)

Short-term debt US$m (1,368) (1,352) (3,461) (1,094) (2,191) (3,519) (3,531) (3,866) (2,793) (953)

Other US$m (3,248) (4,173) (6,243) (4,774) (4,384) (6,496) (6,046) (4,480) (4,480) (4,480)





Other US$m (6,152) (6,888) (8,137) (8,209) (8,588) (10,332) (10,501) (9,339) (8,766) (8,186)



Net assets US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 72,513 87,231 107,167

Equity


Minority interests US$m 237 251 708 757 804 993 1,215 1,243 1,243 1,243

Total equity US$m 24,455 29,918 39,043 40,711 49,329 57,755 67,085 72,513 87,231 107,167




AND LOVE THE ORE

Exhibit 423 BHP Cash Flow Statement


Cash Flow


EBITDA 17,447 21,284 27,757 16,031 24,790 36,855 30,324 25,614 38,119 45,903

Change in working capital US$m - - (3,744) 3,931 (1,719) 28 968 (4,058) (1,693) (1,580)

Other US$m (5,432) (5,329) 1,135 5,220 (825) 198 2,146 1,984 (662) (662)



Royalty-related tax paid US$m (659) (554) (885) (906) (576) (607) (1,015) (1,224) (2,319) (2,944)

Net Interest paid US$m (378) (380) (630) (314) (421) (455) (588) (1,192) (1,586) (1,309)

Other US$m 2,671 4,257 51 30 20 12 25 10 - -



Acquisitions & investments US$m (1,362) (1,667) (1,686) (1,710) (1,048) (5,515) (13,967) (567) - -

Disposals US$m 1,089 408 180 277 386 198 310 2,413 - -

Other US$m - - - (126) - - 6 - - -


(Repayment)/receipt of debt US$m (1,101) 1,382 (750) 3,575 (485) (577) 8,827 7,257 (2,146) (3,679)

Dividends paid to company shareholders US$m (1,936) (2,271) (3,135) (4,563) (4,618) (5,054) (5,877) (6,098) (5,114) (5,257)

Dividends paid to minority shareholders US$m (190) (68) (115) (406) (277) (90) (56) (53) (158) (183)

Shares issued US$m 34 22 - - - - - - - -

Shares bought back US$m (2,215) (5,906) (3,365) (169) (274) (10,329) (507) (348) - -

Other financing activities US$m (4) (2) 366 383 347 32 122 61 - -


Effects of exchange rate on cash US$m 1 11 21 26 26 27 (56) (1) - -

Change in cash and equivalents US$m (426) 1,140 1,775 6,658 1,624 (2,375) (5,035) (1,151) 6,279 10,915

Closing cash and equivalents US$m 776 1,937 4,237 10,833 12,456 10,084 4,781 3,748 10,028 20,943

FCF to firm US$m 2,965 5,354 10,838 9,655 7,968 19,376 6,726 (2,633) 15,284 21,344

FCF per share US$/share .50 .95 1.95 1.74 1.43 3.64 1.26 (.49) 2.87 4.01

(FCF = Operating cash flow less tax paid, less capex)




Exhibit 424 Vale Income Statement


Income Statement


Production

Iron Ore kt 264,153 303,161 301,698 237,953 307,793 322,597 319,961 272,398 297,995 329,993

Manganese Ore kt 2,243 1,333 2,384 1,656 1,842 2,557 2,363 2,008 2,008 2,008

Ferroalloys kt 534 543 473 224 451 436 392 128 128 128

Metallurgical Coal kt - 1,764 2,810 2,528 3,059 2,767 5,083 7,992 11,695 12,465

Thermal Coal kt - 440 1,286 2,894 3,832 4,505 3,419 2,616 4,213 4,543

Nickel kt 232 249 275 187 178 238 235 285 310 314

Copper kt 265 284 312 197 208 300 290 405 469 558

Potash kt 733 671 607 716 662 625 549 480 480 480

Prices

Iron Ore US$/t 37 42 64 55 100 139 108 122 129 132

Manganese Ore US$/t 117 155 639 246 347 271 222 241 267 287

Ferro Manganese US$/t 828 1,357 2,892 1,291 1,374 1,302 1,205 1,130 1,264 1,398


Thermal Coal US$/t 51 61 123 60 77 111 84 77 90 104

Nickel US$/t 24,141 37,183 21,145 14,665 21,813 22,866 17,533 16,517 17,890 19,412

Copper US$/t 6,738 7,133 6,961 5,148 7,533 8,813 7,948 7,825 9,049 10,352

Phosphates US$/t 258 423 967 318 499 628 635 667 742 799

Revenue

Bulk Materials US$m 12,569 15,434 23,553 15,071 34,478 46,904 35,662 33,222 39,305 44,892

Base Metals US$m 5,962 15,313 11,764 6,679 8,200 9,627 7,133 8,720 11,087 13,267

Fertilizers US$m 143 178 295 413 1,846 3,547 3,777 4,790 5,629 6,060

Logistics US$m 1,376 1,525 1,607 1,104 1,465 1,727 1,644 889 906 924

Other US$m 313 665 1,290 672 492 541 537 738 753 767


VAT US$m (716) (873) (1,083) (628) (1,188) (1,399) (1,059) (3,940) (4,514) (5,095)

Net operating revenue US$m 19,647 32,242 37,426 23,311 45,293 60,947 47,694 44,419 53,165 60,815


Depreciation, depletion and amortization US$m (998) (2,190) (2,807) (2,722) (3,260) (4,122) (4,396) (4,753) (5,427) (5,797)


Financial income US$m 327 295 602 381 290 718 401 2,281 1,653 2,642

Interest expense US$m (1,338) (2,482) (1,765) (1,558) (2,646) (2,465) (2,414) (3,919) (3,981) (3,884)

Other financing charges US$m 883 3,449 (448) 2,203 975 (1,566) (1,788) - - -

Gain on sale of investments US$m - 320 777 (870) 40 - 1,513 (491) - -

PBT US$m 7,504 14,772 14,864 6,213 20,354 25,287 11,449 15,240 18,476 25,376

Income tax expense US$m (1,432) (3,201) (535) (2,100) (3,705) (5,282) 833 (4,940) (6,188) (7,985)

Equity results in affiliates US$m 710 595 794 433 987 1,135 (1,001) 726 861 1,102

Loss from discontinued operations US$m - - - - (143) - - - - -

Tax rate % 19.1% 21.7% 3.6% 33.8% 18.2% 20.9% -7.3% 32.4% 33.5% 31.5%

PAT US$m 6,782 12,166 15,123 4,546 17,493 21,140 11,281 11,026 13,149 18,492

Minority interests US$m (579) (802) (258) (107) (189) 233 257 39 (30) (39)


Reported EPS - Preferred US$/share 2.14 2.42 2.53 0.97 3.23 4.37 1.06 2.28 2.55 3.58

Reported EPS - Common US$/share 2.18 2.42 2.53 0.97 3.22 4.34 1.07 2.28 2.55 3.58

Reported EPS - Pref linked to convertibles US$/share 0.00 2.20 4.13 1.73 4.95 6.37 0.00 0.00 0.00 0.00

Reported EPS - Common linked to convertibles US$/share 0.00 2.36 4.31 2.26 7.44 8.15 0.00 0.00 0.00 0.00

DPS US$/share .80 .56 .56 .53 .57 1.73 1.17 .33 .51 .72

Preferred shares outstanding m 1,181 1,889 1,985 2,031 2,025 1,964 1,944 1,968 1,968 1,968

Common shares outstanding m 1,840 2,943 3,094 3,182 3,209 3,187 3,176 3,186 3,186 3,186

Treasury Prefs Shares Linked to Mandatory Convertible Notes m - 23 30 66 47 47 - - - -

Treasury Common Shares Linked to Mandatory Convertible Notes m - 42 57 70 18 18 - - - -



AND LOVE THE ORE

Exhibit 425 Vale Balance Sheet


Balance Sheet


Non-current assets


PP&E US$m 38,007 54,625 48,454 67,637 83,096 88,895 90,744 95,975 105,882 112,684


Other US$m 3,142 3,999 3,119 5,277 5,164 5,843 7,376 7,948 7,948 7,948


Current assets

Inventories US$m 3,493 3,859 3,896 3,196 4,298 5,251 5,052 6,117 7,488 7,568

Receivables US$m 3,604 3,952 3,204 3,120 8,211 8,505 6,795 7,121 8,522 9,465

Cash and equivalents US$m 4,448 1,046 10,331 7,293 7,584 3,531 5,832 4,162 4,338 10,404

Other US$m 1,395 2,523 5,807 7,685 11,698 4,449 5,218 5,299 5,299 5,299


Total assets US$m 60,926 76,717 79,992 102,279 129,139 128,728 131,478 142,887 156,601 171,595

Current liabilities

Payables US$m (2,382) (2,430) (2,261) (2,309) (3,558) (4,814) (4,529) (4,646) (5,687) (5,748)

Short-term debt US$m (1,434) (1,416) (633) (2,963) (2,962) (1,517) (3,468) (3,563) (3,534) (2,523)

Other US$m (3,496) (6,237) (4,343) (3,909) (8,240) (4,712) (4,407) (3,871) (3,871) (3,871)





Other US$m (5,481) (7,470) (5,789) (6,196) (10,978) (9,043) (10,721) (14,002) (16,209) (18,362)



Net assets US$m 22,484 35,831 45,047 60,497 72,441 80,114 76,363 84,459 94,984 109,786

Equity



Total equity US$m 22,484 35,831 45,047 60,497 72,441 80,114 76,363 84,459 94,984 109,786





Exhibit 426 Vale Cash Flow Statement


Cash Flow


EBITDA 8,620 15,380 18,505 8,779 24,955 32,722 18,133 22,523 26,230 32,415

Change in working capital US$m 288 1,060 (229) 1,426 (3,083) (803) 1,621 (2,137) (1,730) (963)

Other US$m (636) (604) 2,021 (624) 871 1,317 (3,081) 691 - -


Income tax paid US$m (541) (3,284) (2,867) (1,331) (1,972) (7,293) 1,238 (4,833) (6,188) (7,985)

Net interest paid US$m (498) (1,338) (1,266) (1,114) (1,102) (1,146) (1,316) (94) (121) 911



Acquisitions & investments US$m (13,201) (2,926) - (1,952) (6,252) - - (182) - -

Disposals US$m 886 1,042 134 448 - 1,081 974 700 - -

Other US$m (208) (471) (2,563) (3,559) 1,715 625 (544) (345) - -


(Repayment)/receipt of debt US$m 15,011 (2,620) 559 2,471 1,907 (1,135) 7,621 332 (59) (2,022)

Dividends paid to shareholders US$m (1,319) (1,875) (2,850) (2,724) (3,000) (9,000) (6,000) (1,726) (2,624) (3,691)

Dividends paid to minority interests US$m (46) (714) (143) (47) (140) (100) (45) - - -

Shares issued US$m - - - - - (1,134) (411) - - -

Shares bought back US$m (301) - (752) 925 (1,996) (3,002) - - - -

Other financing activities US$m - - 12,190 - 660 - - - - -


Effects of exchange rate on cash US$m (217) (199) (5,432) 2,360 375 (109) (112) 11 - -

Change in cash and equivalents US$m 3,407 (3,402) 8,335 (3,038) 291 (4,052) 2,301 (1,670) 176 6,066

Closing cash and equivalents US$m 4,448 1,046 10,331 7,293 7,584 3,531 5,832 4,162 4,338 10,404

FCF to firm US$m 3,300 5,699 8,458 154 8,124 9,868 2,134 (365) 2,979 10,868

FCF per share US$/share 1.09 1.16 1.64 .03 1.53 1.89 .42 (.07) .58 2.11




AND LOVE THE ORE

Exhibit 427 Anglo American Income Statement


Income Statement


Production

Platinum koz 2,817 2,474 2,387 2,452 2,570 2,530 2,379 2,262 2,386 2,378

Palladium koz 1,539 1,390 1,319 1,361 1,449 1,431 1,396 1,356 1,442 1,448

Diamonds k carats 23,011 23,001 21,659 11,070 14,849 14,098 18,314 24,524 24,524 24,524

Copper kt 680 663 679 677 633 610 670 696 703 735

Nickel kt 26 26 20 20 20 29 39 28 38 41

Iron Ore kt 31,109 32,400 37,411 44,596 47,482 46,089 49,137 49,040 46,008 59,234

Thermal Coal - Domestic kt 43,390 42,521 44,412 51,080 49,923 47,592 46,939 48,798 48,798 48,798

Thermal Coal - Export kt 33,762 35,212 33,771 25,960 26,808 27,217 28,076 28,486 29,484 30,816

Metallurgical Coal kt 10,964 11,289 14,889 12,934 17,034 15,243 18,864 19,201 19,201 21,071

Prices

Platinum US$/oz 1,142 1,301 1,576 1,200 1,611 1,722 1,552 1,598 1,733 1,859

Palladium US$/oz 320 355 358 267 526 716 605 741 766 785

Copper US$/t 6,725 7,124 6,969 5,142 7,531 8,822 7,952 7,830 9,049 10,352

Nickel US$/t 24,102 37,147 21,155 14,655 21,799 22,896 17,544 16,507 17,890 19,412

Iron Ore Lump US$/t 54 60 105 80 125 173 127 139 147 151


Thermal Coal US$/t 49 66 129 72 99 121 94 92 104 116


Revenue

Platinum US$m 5,861 6,789 6,327 4,535 6,602 7,359 5,489 5,610 6,336 6,727

Diamonds US$m 3,148 3,076 3,096 1,728 2,644 3,320 4,028 5,709 6,314 6,426

Copper US$m 4,537 4,507 3,907 3,967 4,877 5,144 5,122 5,346 6,238 7,467

Nickel US$m 553 878 408 348 426 488 336 444 664 776

Iron Ore and Manganese US$m 2,684 2,300 4,099 3,419 6,612 8,124 6,730 7,497 8,136 10,472

Metallurgical Coal US$m 1,398 1,389 3,119 2,239 3,377 4,347 3,889 3,521 3,993 4,699

Thermal Coal US$m 1,935 2,165 3,157 2,490 2,866 3,722 3,447 3,122 3,721 4,401

Other Mining & Industrial US$m 18,521 9,455 8,851 5,908 5,520 4,039 3,739 3,604 3,604 3,604

Other US$m - - - 3 5 5 5 4 4 4


Less associates' revenue US$m (5,565) (5,089) (6,653) (3,779) (4,969) (5,968) (4,024) (2,403) (2,817) (3,291)



Associate income US$m 582 640 1,303 318 845 978 493 155 273 395


Net interest US$m (165) (137) (452) (273) (244) (20) (288) (456) (378) (284)

PBT (underlying) US$m 9,159 9,021 8,832 4,422 9,109 10,626 5,610 5,626 7,228 9,727

Income tax expense US$m (2,763) (2,676) (2,545) (1,305) (2,699) (2,741) (1,488) (1,519) (1,952) (2,626)

Tax rate % 30.2% 29.7% 28.8% 29.5% 29.6% 25.8% 26.5% 27.0% 27.0% 27.0%





EPS incl special items USD/share 4.23 6.00 4.34 2.02 5.43 5.09 (1.17) 1.90 2.57 3.75

Diluted EPS incl special items USD/share - - - 1.98 5.18 4.89 (1.14) 1.90 2.57 3.74

DPS USD/share .67 .80 .92 - .25 .68 .78 .70 .45 .59

Shares outstanding m 1,461 1,218 1,201 1,202 1,206 1,211 1,278 1,278 1,278 1,278




Exhibit 428 Anglo American Balance Sheet


Balance Sheet


Non-current assets


PP&E US$m 23,498 23,534 29,545 35,198 39,810 40,549 45,089 47,662 49,713 48,395


Other US$m 3,039 5,611 3,995 3,991 4,952 5,029 5,449 5,710 5,937 6,228


Current assets

Inventories US$m 2,974 2,344 2,702 3,212 3,604 3,517 5,005 4,979 5,464 5,729

Receivables US$m 5,312 3,731 2,929 3,351 3,731 3,674 3,275 4,019 4,475 5,157

Cash and equivalents US$m 3,004 3,129 2,771 3,269 6,401 11,732 9,094 10,103 10,220 13,519

Other US$m 1,742 1,516 1,178 1,199 942 379 3,823 673 673 673


Total assets US$m 46,483 44,762 49,738 56,308 66,656 72,442 79,369 80,857 84,330 87,746

Current liabilities

Payables US$m (5,040) (3,950) (4,770) (4,395) (4,950) (5,098) (4,536) (6,368) (6,989) (7,328)

Short-term debt US$m (2,028) (5,895) (6,784) (1,499) (1,535) (1,018) (2,604) (1,492) (951) -

Other US$m (1,731) (1,635) (1,570) (851) (1,397) (2,062) (1,663) (1,681) (1,704) (1,735)





Other US$m (2,650) (1,898) (3,092) (3,486) (3,258) (3,490) (5,560) (4,817) (4,993) (5,169)



Net assets US$m 27,127 24,330 21,756 28,069 37,971 43,189 43,787 45,888 49,270 54,100

Equity

Shareholders' equity US$m 24,271 22,461 20,221 26,121 34,239 39,092 37,657 39,170 41,856 45,878


Total equity US$m 27,127 24,330 21,756 28,069 37,971 43,189 43,787 45,888 49,270 54,100

Net debt/(Net cash) US$m 3,051 4,782 11,348 11,328 7,443 1,374 8,828 5,966 4,947 (544)



AND LOVE THE ORE

Exhibit 429 Anglo American Cash Flow Statement


Cash Flow


EBITDA 12,197 11,171 11,790 6,930 11,983 13,463 8,815 10,052 11,833 15,590

Share of associates operating profit US$m (1,090) (1,072) (2,104) (580) (1,255) (1,427) (759) (496) (823) (1,188)

Share of associates D&A US$m (329) (183) (253) (248) (301) (286) (233) (161) (173) (177)

Change in working capital US$m (532) (688) (23) (910) (380) (159) (527) 984 (400) (727)

Other US$m (189) 611 112 (288) (123) (93) (275) 176 176 176


Dividends received US$m 288 311 659 639 285 403 340 110 169 230


Other US$m - - - - - - - - - -



Acquisitions & investments US$m (344) (1,933) (7,246) (384) (653) (63) (4,946) - - -

Disposals US$m 1,766 822 1,554 2,818 2,863 610 166 - - -

Other US$m 459 2,781 (912) (50) 600 803 566 2,457 226 226


(Repayment)/receipt of debt US$m 583 3,121 6,613 (371) (1,144) (297) 4,886 (1,854) (901) (2,192)

Dividends paid to minorities US$m (383) (728) (796) (472) (617) (1,404) (1,267) (1,092) (1,293) (1,501)

Dividends paid to shareholders US$m (2,888) (1,538) (1,550) - (302) (818) (970) (895) (579) (749)

Interest paid US$m (426) (483) (741) (741) (837) (807) (775) (845) (783) (714)

Shares issued US$m - - - - - - - - - -

Shares bought back US$m - - - - - - - - - -

Other financing activities US$m (3,561) (5,341) 16 (35) 500 4,800 76 (19) (22) (21)


Change in cash and equivalents US$m (170) 34 (200) 245 2,857 5,983 (3,049) 1,009 117 3,299

Closing cash and equivalents US$m 3,004 3,129 2,771 3,269 6,401 11,732 9,094 10,103 10,220 13,519

FCF to firm US$m 4,624 3,333 2,862 (520) 2,447 3,159 (785) 3,257 3,470 8,250

FCF per share US$/share 3.16 2.74 2.38 (.43) 2.03 2.61 (.61) 2.55 2.72 6.46





Exhibit 430 Glencore Xstrata Pro Forma Income Statement (Including Synergies)


Income Statement


Sales Volume

Zinc Own Feed kt 1,795 1,814 1,761 1,668 1,569 1,601 1,955 2,028

Copper Own Feed kt 1,198 1,206 1,199 1,241 1,161 1,446 1,578 1,801

Nickel & Ferronickel Own Feed kt 91 80 85 100 104 123 145 167

Coal Mt 107 104 100 105 137 153 162 176

Ferrochrome kt 1,126 786 1,165 1,021 918 990 1,204 1,204

Platinum koz 138 133 118 92 76 76 148 244

Oil Own Feed Mstb 17 9 - 1 5 5 6 6

Oil - Share of RussNeft Investment Mstb - - 3 5 6 6 6 6

Prices

Zinc US$/t 1,886 1,656 2,158 2,195 1,949 2,036 2,237 2,387

Copper US$/t 6,969 5,142 7,531 8,822 7,952 7,830 9,049 10,352

Nickel US$/t 21,155 14,655 21,799 22,896 17,544 16,507 17,890 19,412

Thermal Coal US$/t 129 72 99 121 94 92 104 116

Platinum US$/oz 1,576 1,200 1,611 1,722 1,552 1,598 1,733 1,859

Brent US$/bbl 99 62 80 111 110 121 135 145

Revenue

Zinc US$m 5,671 5,445 6,678 7,047 7,117 6,913 8,221 8,878

Copper US$m 14,203 11,554 17,435 19,213 16,191 18,209 21,742 26,687

Nickel US$m 3,754 2,500 3,451 3,872 3,306 3,346 4,134 5,010

Alumina & Aluminum US$m 263 235 422 520 410 454 475 532

Coal US$m 9,175 7,816 9,035 11,648 12,205 12,576 15,216 18,183

Oil & Oil Products US$m 1,672 534 253 642 1,640 1,828 2,206 2,457

Agricultural Products US$m 1,337 1,789 2,180 3,359 2,935 3,100 3,153 3,211

Alloys US$m 2,002 1,305 1,894 1,689 1,413 1,434 1,997 2,380

Marketing - Metals & Mining US$m 34,940 30,971 38,639 44,067 50,834 48,703 52,789 57,291

Marketing - Energy Products US$m 87,310 54,041 80,061 104,775 118,192 129,598 145,034 155,637

Marketing - Agricultural Products US$m 12,057 6,793 8,238 13,744 14,155 12,284 12,356 12,434

Other US$m 235 114 153 222 358 358 358 358


Operating costs US$m (161,545) (116,778) (157,970) (199,700) (220,589) (230,466) (255,535) (274,731)


Financial income US$m 490 674 433 476 579 437 532 809

Financial expense US$m (1,987) (1,608) (1,837) (1,638) (1,808) (1,688) (1,628) (1,452)

PBT US$m 10,327 6,135 9,815 10,686 7,688 7,835 11,799 18,432

Income tax expense US$m (2,051) (1,402) (2,194) (2,033) (1,065) (1,485) (2,499) (4,209)

Tax rate % 19.9% 22.8% 22.4% 19.0% 13.8% 19.0% 21.2% 22.8%

PAT US$m 8,276 4,733 7,621 8,653 6,623 6,350 9,299 14,223

Minority interests US$m (339) (296) (475) (456) (497) (672) (951) (1,286)


Reported EPS US$/share .59 .33 .54 .61 .46 .43 .63 .97

DPS US$/share .06 .06 .06 .09 .14 .14 .19 .23

Post-acquisition shares out m 13,326 13,326 13,326 13,326 13,326 13,326 13,326 13,326



AND LOVE THE ORE

Exhibit 431 Glencore Xstrata Pro Forma Balance Sheet (Including Synergies)


Balance Sheet

December year-end 2008 A 2009 A 2010 A 2011 A 2012 E 2013 E 2014 E 2015 E

Non-current assets

Intangible assets US$m 6,714 6,238 6,216 6,254 6,850 6,850 6,850 6,850

PP&E US$m 43,000 46,242 57,972 66,093 78,146 78,191 76,858 74,355

Other US$m 9,887 11,027 11,677 11,836 10,930 10,941 11,101 11,562

Total non-current assets US$m 59,601 63,507 75,865 84,183 95,926 95,983 94,810 92,767

Current assets

Inventories US$m 11,378 19,643 22,156 22,371 20,435 22,515 24,824 26,744

Receivables US$m 15,897 18,495 23,457 25,637 25,110 28,230 31,015 34,143

Cash and equivalents US$m 1,982 2,037 3,185 3,253 2,130 2,702 4,736 9,903

Other US$m 14,238 10,967 7,135 5,845 6,024 6,024 6,024 6,024

Total current assets US$m 43,495 51,142 55,933 57,106 53,699 59,471 66,600 76,814

Total assets US$m 103,096 114,649 131,798 141,289 149,625 155,454 161,409 169,581

Current liabilities

Payables US$m (14,847) (15,179) (20,775) (23,262) (21,878) (24,217) (26,574) (28,601)

Short-term debt US$m (6,039) (10,172) (14,683) (9,790) (8,808) (8,215) (6,505) (3,800)

Other US$m (14,880) (10,137) (10,233) (6,886) (6,935) (6,941) (6,952) (6,972)

Total current liabilities US$m (35,766) (35,488) (45,691) (39,938) (37,621) (39,373) (40,031) (39,374)


Long-term debt US$m (29,408) (29,655) (25,405) (28,648) (28,766) (28,173) (26,463) (23,758)

Other US$m (10,741) (11,798) (13,651) (14,375) (17,194) (17,594) (18,062) (18,665)

Total non-current liabilities US$m (40,149) (41,453) (39,056) (43,023) (45,960) (45,767) (44,525) (42,423)


Net assets US$m 27,181 37,412 46,850 58,328 66,044 70,314 76,853 87,785

Equity

Shareholders equity US$m 24,639 34,517 42,194 53,221 59,753 63,585 69,397 79,288

Minority interests US$m 2,542 2,895 4,656 5,107 6,290 6,729 7,457 8,496

Total equity US$m 27,181 37,412 46,850 58,328 66,044 70,314 76,853 87,785





Exhibit 432 Glencore Xstrata Pro Forma Cash Flow Statement (Including Synergies)


Cash Flow


EBITDA 14,638 9,949 14,708 15,878 13,784 15,902 20,203 26,252

Change in working capital US$m 2,682 (6,970) (3,523) (2,691) 1,754 (2,891) (2,812) (3,114)

Other US$m 2 891 750 (175) (75) 187 384 776


Income tax paid US$m (2,170) (1,158) (1,854) (1,557) (1,965) (1,483) (2,585) (4,515)

Net interest paid US$m (1,449) (1,085) (1,268) (1,226) (1,315) (1,048) (1,002) (844)

Other US$m 850 585 345 403 431 427 413 382


Capital expenditure US$m (6,645) (4,665) (7,709) (10,844) (11,224) (7,006) (6,119) (4,817)

Acquisitions & investments US$m (6,909) (390) (1,238) (1,478) 2,143 - - -

Disposals US$m 352 1,161 652 379 378 1,206 144 10

Other US$m 74 (2,948) 382 (301) (168) (895) 415 1,052


(Repayment)/receipt of debt US$m 2,380 (769) 2,101 (1,466) (2,383) (1,186) (3,420) (5,409)

Dividends paid to company shareholders US$m (826) (782) (817) (1,149) (1,887) (1,872) (2,584) (3,061)

Dividends paid to minority shareholders US$m (522) - (407) (1,002) (1,209) (365) (613) (1,188)

Shares issued US$m - 5,667 - 7,616 - - - -

Shares bought back US$m - - - - - - - -

Other financing activities US$m (1,874) 615 (1,055) (1,586) 385 (403) (389) (358)


Effects of exchange rate on cash US$m - - - - - - - -

Change in cash and equivalents US$m 583 101 1,067 801 (1,350) 572 2,034 5,167

Closing cash and equivalents, Dec 31 US$m 1,982 2,037 3,185 3,253 2,130 2,702 4,736 9,903

FCF to firm US$m 8,507 (1,953) 2,372 611 2,275 4,709 9,071 14,582

FCF per share US$/share .64 (.15) .18 .05 .17 .35 .68 1.09




AND LOVE THE ORE

Exhibit 433 Glencore Xstrata Pro Forma Income Statement (Excluding Synergies)


Income Statement


Sales Volume

Zinc Own Feed kt 1,795 1,814 1,761 1,668 1,569 1,601 1,955 2,028

Copper Own Feed kt 1,198 1,206 1,199 1,241 1,161 1,446 1,578 1,801

Nickel & Ferronickel Own Feed kt 91 80 85 100 104 123 145 167

Coal Mt 107 104 100 105 137 153 162 176

Ferrochrome kt 1,126 786 1,165 1,021 918 990 1,204 1,204

Platinum koz 138 133 118 92 76 76 148 244

Oil Own Feed Mstb 17 9 - 1 5 5 6 6

Oil - Share of RussNeft Investment Mstb - - 3 5 6 6 6 6

Prices

Zinc US$/t 1,886 1,656 2,158 2,195 1,949 2,036 2,237 2,387

Copper US$/t 6,969 5,142 7,531 8,822 7,952 7,830 9,049 10,352

Nickel US$/t 21,155 14,655 21,799 22,896 17,544 16,507 17,890 19,412

Thermal Coal US$/t 129 72 99 121 94 92 104 116

Platinum US$/oz 1,576 1,200 1,611 1,722 1,552 1,598 1,733 1,859

Brent US$/bbl 99 62 80 111 110 121 135 145

Revenue

Zinc US$m 5,671 5,445 6,678 7,047 7,117 6,913 8,221 8,878

Copper US$m 14,203 11,554 17,435 19,213 16,191 18,209 21,742 26,687

Nickel US$m 3,754 2,500 3,451 3,872 3,306 3,346 4,134 5,010

Alumina & Aluminum US$m 263 235 422 520 410 454 475 532

Coal US$m 9,175 7,816 9,035 11,648 12,205 12,576 15,216 18,183

Oil & Oil Products US$m 1,672 534 253 642 1,640 1,828 2,206 2,457

Agricultural Products US$m 1,337 1,789 2,180 3,359 2,935 3,100 3,153 3,211

Alloys US$m 2,002 1,305 1,894 1,689 1,413 1,434 1,997 2,380

Marketing - Metals & Mining US$m 34,565 30,221 37,889 43,317 50,084 47,953 52,039 56,541

Marketing - Energy Products US$m 87,310 54,041 80,061 104,775 118,192 129,598 145,034 155,637

Marketing - Agricultural Products US$m 12,057 6,793 8,238 13,744 14,155 12,284 12,356 12,434

Other US$m 235 114 153 222 358 358 358 358


Operating costs US$m (161,315) (116,174) (157,366) (199,096) (219,984) (229,862) (254,931) (274,127)


Financial income US$m 490 674 433 476 579 437 532 809

Financial expense US$m (1,987) (1,608) (1,837) (1,638) (1,808) (1,688) (1,628) (1,452)

PBT US$m 9,432 5,239 8,919 9,790 6,792 6,939 10,903 17,536

Income tax expense US$m (1,902) (1,231) (2,016) (1,876) (956) (1,340) (2,326) (4,012)

Tax rate % 20.2% 23.5% 22.6% 19.2% 14.1% 19.3% 21.3% 22.9%

PAT US$m 7,530 4,008 6,903 7,914 5,836 5,599 8,577 13,524

Minority interests US$m (339) (296) (475) (456) (497) (672) (951) (1,286)


Reported EPS US$/share .54 .28 .48 .56 .40 .37 .57 .92

DPS US$/share .02 .02 .02 .04 .09 .10 .15 .19

Post-acquisition shares out m 13,326 13,326 13,326 13,326 13,326 13,326 13,326 13,326




Exhibit 434 Glencore Xstrata Pro Forma Balance Sheet (Excluding Synergies)


Balance Sheet


Non-current assets

Intangible assets US$m 8,898 8,422 8,400 8,438 9,034 9,034 9,034 9,034

PP&E US$m 43,000 46,242 57,972 66,093 78,146 78,191 76,858 74,355

Other US$m 9,887 11,027 11,677 11,836 10,930 10,941 11,101 11,562

Total non-current assets US$m 61,785 65,691 78,049 86,367 98,110 98,167 96,994 94,951

Current assets

Inventories US$m 11,378 19,643 22,156 22,371 20,435 22,515 24,824 26,744

Receivables US$m 15,897 18,495 23,457 25,637 25,110 28,230 31,015 34,143

Cash and equivalents US$m 1,982 2,037 3,185 3,253 2,130 2,702 4,736 9,903

Other US$m 14,238 10,967 7,135 5,845 6,024 6,024 6,024 6,024

Total current assets US$m 43,495 51,142 55,933 57,106 53,699 59,471 66,600 76,814

Total assets US$m 105,280 116,833 133,982 143,473 151,808 157,638 163,593 171,765

Current liabilities

Payables US$m (14,847) (15,179) (20,775) (23,262) (21,878) (24,217) (26,574) (28,601)

Short-term debt US$m (6,039) (10,172) (14,683) (9,790) (8,808) (8,215) (6,505) (3,800)

Other US$m (14,880) (10,137) (10,233) (6,886) (6,935) (6,941) (6,952) (6,972)

Total current liabilities US$m (35,766) (35,488) (45,691) (39,938) (37,621) (39,373) (40,031) (39,374)


Long-term debt US$m (29,408) (29,655) (25,405) (28,648) (28,766) (28,173) (26,463) (23,758)

Other US$m (10,741) (11,798) (13,651) (14,375) (17,194) (17,594) (18,062) (18,665)

Total non-current liabilities US$m (40,149) (41,453) (39,056) (43,023) (45,960) (45,767) (44,525) (42,423)


Net assets US$m 29,365 39,596 49,034 60,512 68,227 72,498 79,037 89,968

Equity

Shareholders equity US$m 26,823 36,701 44,378 55,405 61,937 65,769 71,580 81,472

Minority interests US$m 2,542 2,895 4,656 5,107 6,290 6,729 7,457 8,496

Total equity US$m 29,365 39,596 49,034 60,512 68,227 72,498 79,037 89,968




AND LOVE THE ORE

Exhibit 435 Glencore Xstrata Pro Forma Cash Flow (Excluding Synergies)


Cash Flow


EBITDA 13,888 9,199 13,958 15,128 13,034 15,152 19,453 25,502

Change in working capital US$m 2,682 (6,970) (3,523) (2,691) 1,754 (2,891) (2,812) (3,114)

Other US$m 2 891 750 (175) (75) 187 384 776


Income tax paid US$m (2,021) (987) (1,676) (1,400) (1,857) (1,339) (2,412) (4,318)

Net interest paid US$m (1,449) (1,085) (1,268) (1,226) (1,315) (1,048) (1,002) (844)

Other US$m 850 585 345 403 431 427 413 382


Capital expenditure US$m (6,645) (4,665) (7,709) (10,844) (11,224) (7,006) (6,119) (4,817)

Acquisitions & investments US$m (6,909) (390) (1,238) (1,478) 2,143 - - -

Disposals US$m 352 1,161 652 379 378 1,206 144 10

Other US$m 74 (2,948) 382 (301) (168) (895) 415 1,052


(Repayment)/receipt of debt US$m 2,380 (769) 2,101 (1,466) (2,383) (1,186) (3,420) (5,409)

Dividends paid to company shareholders US$m (225) (203) (245) (556) (1,245) (1,267) (2,007) (2,508)

Dividends paid to minority shareholders US$m (522) - (407) (1,002) (1,209) (365) (613) (1,188)

Shares issued US$m - 5,667 - 7,616 - - - -

Shares bought back US$m - - - - - - - -

Other financing activities US$m (1,874) 615 (1,055) (1,586) 385 (403) (389) (358)

Net cash flow from financing US$m (241) 5,310 394 3,006 (4,452) (3,221) (6,429) (9,462)

Effects of exchange rate on cash US$m - - - - - - - -

Change in cash and equivalents US$m 583 101 1,067 801 (1,350) 572 2,034 5,167

Closing cash and equivalents, Dec 31 US$m 1,982 2,037 3,185 3,253 2,130 2,702 4,736 9,903

FCF to firm US$m 7,906 (2,532) 1,800 18 1,633 4,103 8,494 14,029

FCF per share US$/share .59 (.19) .14 .00 .12 .31 .64 1.05





Index of Exhibits

1 Financial Overview 4

2 In 2000, China Was a Clear Outlier — the World's Largest Country Set to Age Rapidly and Yet Still at a Pre-Industrial Output Level 5

3 The Demographic Challenges Facing China Stand Behind the Requirement to Accelerate the Capital Stock Formation 5

4 China Accounted for ~16% of 2012 Global Iron Ore Production (ROE Equivalent); Our Analysis Shows That ~60% of the Chinese Production Consisted Primarily of "Missing" Chinese High-Cost Mines 7

5 Productivity Uplift of the Iron Ore Imports Into China Has Enabled It to Overtake the U.S. in Terms of Effective Productivity in 2003 7

6 The Value Destructive Effect of Excessive Volume Growth Can Be Seen Once the Impact of Supply Elasticity Is Considered 8

7 It Is the Elasticity of Supply That Inflects Any Commodity Cost Curve; the Higher the Elasticity, the Higher the Margin and the Lower the Degree of Consolidation Required to Generate Superior Returns 8

8 Our Three Scenarios Set the Widest Plausible Band, in Our View, Around Implied Consensus Expectations; They Range from More Bullish Than the IMF's Current Forecast, Through Official Chinese GDP Targets, to a Bear Scenario in Which No Supply Is Removed from the Market Despite Prices Collapsing 9

9 Our Iron Ore Price Forecast Is Consistent With the IMF's Current GDP Forecast of 7.8% in 2013 Rising to 8.5% in 2016 and Average Increase in Supply of 88Mtpa 2013-2016 9

10 Mining Costs and Revenue Exhibit a High Degree of Correlation 10

11 Currency Appreciation Drives a Significant Part of the Cost to Price Relationship, Particularly for the Australian Producers 10

12 There Is a Significant Range of Variation Around the Benchmark 62% CIF Price Point, and Even After Adjusting for Fe Grade, Poor-Quality Ore Suffers a Significant Discount 11

13 We Believe That a Significant Part of the Slowdown in Chinese Steel Growth Resulted from the Budgetary Constraint Imposed by High Commodity Prices 11

14 There Is a Clear Relationship Between the Aging of a Population and Its Wealth…Usually, Birth Rates Fall as Societies Get Rich, With Aging Population Supported by the Increased Productivity Associated With Industrialization 16

15 In 1910, the U.S. Began Transitioning from an Agricultural Economy… 17

16 …Toward One Driven by Services 17

17 We Believe That the Increases in Labor Productivity, Occasioned by the Increased Intensity of Capital Stock (Machinery, Railroads, etc.), Enabled This Transition 17



AND LOVE THE ORE

18 In 2000, China Was a Clear Outlier from the Normal Pattern — the World's Largest Country Was Set to Age Rapidly and Yet Was Still at a Pre-industrial Level of Output 18

19 It Was Only in the Year 2000 That China Started Reducing Its Framing Dependence 19

20 1970s Saw the Rise in Construction and Manufacturing Employment Start… 19

21 …Followed in Short Order by the Service Sector 19

22 Today, More People Are Employed in Services Than Manufacturing 19

23 The Pattern in China Is Nearly Identical to That Seen in the U.S.; Capitalization Drives Productivity Increases, Which Then Lead to the Emergence of Services 20

24 Comparing China to the U.S. Suggests That the Structure of the Chinese Economy Today Is Similar to the U.S. in the 1930s, Both in Labor and Capital Stock; However, the Transition Has Been Far More Rapid in China 20

25 The Demographic Challenges Facing China and the Emergence of Lewis Point Stand Behind the Requirement to Accelerate the Capital Stock Formation 21

26 The Urgency to Industrialize the Chinese Economy Is Evident in the Acceleration in Steel Demand in Comparison to the Historical Precedent in the West 22

27 A Very Similar Demographic to China Was Seen in Japan Toward the End of the Last Century; However, by the Time the Japanese Transition Occurred, Industrialization Was Complete and Output Stayed High 22

28 A Rapid Fall in the Child Dependency Ratio Following the "One Child" Policy… 24

29 …Will Eventually Be Offset by the Extraordinary Increase in the Old-Age Dependency Ratio 24

30 The End of the Last Century Marked an Opportunity to Deploy Abundant Labor in China to Develop Its Capital Stock… 24

31 …Mirroring Japan But Scaling Up More Than 10 Fold (127 Million vs. 1,345 Million People) 24

32 1930-1973 Saw Real Commodity Prices Grow as Capital Was Accumulated in Western Economies; Once Demand Went Flat, Productivity Improvements in Mining Drove Real Commodity Prices Down 26

33 The Emergence of a New Phase of Capital Accumulation Driven by China Has Seen Significant Real Commodity Price Appreciation; Our Counter-Consensus View on Commodities Is Driven, in Part, by a Belief That This Will Continue Until the Chinese Capital Stock Accumulation Comes to an End 26

34 Chinese Labor Costs Have Been Instrumental in Changing Global Commodity Cost Curve Structures and, With Them, the Price of Mined Raw Materials; What Happens Once China Moves Past the Lewis Point? 27

35 Across Our Coverage Group, Iron Ore Generated, on Average, 41% of Revenue… 30

36 ...And 65% of EBITDA in 2012 30




37 Iron Ore Currently Accounts for 44% of Rio Tinto's Revenue and Has Been Driven by a 7.6% Production CAGR 30

38 Iron Ore Currently Accounts for 31% of BHP Billiton's Revenue and Has Been Driven by an 8.1% Production CAGR 31

39 Iron Ore Currently Accounts for 70% of Vale's Revenue and Has Been Driven by a 6.9% Production CAGR 31

40 Iron Ore Currently Accounts for 18% of Anglo American's Revenue and Has Been Driven by a 23.7% Production CAGR 32

41 Like Other Countries That Have Industrialized, China's Steel Intensity Has Passed from "Development" Through "Peak" and Is Now in "Decline"; the Steady-State Rate of Consumption Will Be Determined by Whether the Economy, Once Mature, Adopts a More Manufacturing or Services-Oriented Focus 34

42 The "Services and Technology" Pattern (Embodied by the U.S.) Shows the Highest Output/Capital Stock (on a Per-Capita Basis), the "Manufacturing" Pattern (Embodied by Germany) Also Shows an Upward-Sloping Relationship Between Output and Capital Stock; Only the Failed "Inefficient" Pattern (Embodied by the former USSR) Shows a Flat and Negative Relationship Between Output and Embedded Capital Stock 34

43 Our Analysis Derives a Consistent Relationship Between Steel Demand Growth and Overall Level of Economic Activity for China; It Does Not See Signs That the Output Generated by the Embedded Capital Stock Is "Inefficient"; Rather, It Is Increasing in a Manner Consistent With Development Into an Economy Midway Between the "Services and Technology" (Embodied by the U.S.) and the "Manufacturing" (Embodied by Germany) Patterns 35

44 A Central Plank of Our Thesis Rests on the Fact That Chinese Urbanization and Industrialization Are Still Far from Complete; in Terms of Capital Stock Accumulation, Labor Composition and Levels of Output, China Today Resembles the U.S. in the 1930s 36

45 At First Glance, the Old Heuristic of Marginal Cash Costs Appears to Have Broken Down… 38

46 …But the Fact That Our Visibility Into Costs in China Is Incomplete Provides a Better Answer 38

47 China Accounted for ~16% of 2012 Global Iron Ore Production (ROE Equivalent); Our Analysis, Which Completes the Global Cost Curve, Shows That ~60% of the Chinese Production Consisted Primarily of "Missing" Chinese High-Cost Mines 38

48 Our Analysis of the Cash Costs of 149 Green and Brownfield Projects Shows That the Majority Will Not Be Value Accretive at Iron Ore Prices Below $150/t Against a Fully Loaded Discount Rate 38

49 We Embed a Contango Rather Than Backwardation Into Our Short- to Medium-Term Price Forecasts 39

50 Our Iron Ore Forecast Sees Considerable Upside in 2014-16E Relative to Consensus; We Would Be Even More Bullish If We Were Convinced the Western Majors Were Committed to a Higher Degree of Capital Discipline 39

51 Mining in Western Australia in 1961... 43

52 ...And Today 43



AND LOVE THE ORE

53 At Either End of the Industrialization Process, Metal Intensity (Rate of Consumption/Embedding) Is Low; Rather, Metal Is Required to Embed the Capital Stock That Enables the Transition from Primary to Tertiary Forms of Value-Add (or Farmers to Hairdressers) 44

54 The Chinese Economy Has Seen Steel Intensity Peak at ~65kg/$'000 of Output (vs. 40kg/$000 U.S. Peak); Because China's Peak Was More Than 60% Higher Than the U.S., and Because Many Datasets Look at Only the Last ~20 Years of Data, This Has Spawned Fears That the High Rate of Steel Consumption in Recent Years Is Indicative of Overinvestment and That the Chinese Economy Is Unduly (and Inefficiently) Steel Intensive 45

55 Global Freight Rates Have Declined in Real Terms by an Average of 0.9% p.a. Over the Last 260 Years... 46

56 …Leading to the Emergence of a Truly Global Market for Bulk Commodities, Including Iron Ore and Coal; Consequently, Relative Global Competitiveness in the Production of These Commodities Began to Matter 47

57 Before the Advent of Scrap-Based EAF Production, U.S. Iron Ore Production Peaked in the 30 Years Between 1950 and 1980 47

58 The Transition to EAF-Based Production Saw Declines in Both Iron Ore Production and Employment 48

59 Consequently, the U.S. Iron Ore Productivity Peaked at Nearly 9,000 Tons/Man-Year (Although for Much of the U.S. Industrialization, Its Intensity Was Far Lower) 48

60 In Contrast to the U.S., Western Australian Iron Ore Industry Was Always Driven by Exports — First to Japan and Today to China; the Declines in Shipping Costs Facilitated This Greatly 50

61 Superior Geology and Continuing Improvement in Mining and Logistics Technology Have Led to Substantial Increases in Mining Productivity 50

62 In Addition, Low Population Density Renders Iron Ore Mining in Pilbara Hugely Productive... 51

63 …And Provides a Significant Advantage Over the U.S. or China as Mining Jurisdictions 51

64 The Consolidation of the Industry Can Be Seen in Both the Concentration of Employment... 52

65 ...And Production in Western Australia 52

66 Among the Majors, Rio and BHP Have Almost Identical Productivities and Are Almost Twice as Productive as Either FMG or Cliffs 53

67 It Is Also Interesting to Note the Significant Declines in Productivity Seen at the Company Level as Well as in the Industry Overall in Recent Years 53

68 However, the Decline in Productivity in Western Australia Is Due to the Inclusion of Labor That Is Allocated to Building Projects (With a Significant Production Lead Time) vs. Labor Allocated to Operations That Generate Current Volume 54

69 China Has Seen Both Its Steel… 55

70 …And Iron Ore Production Increase 55




71 The Iron Ore Majors (Vale, Rio and BHP) Have Aggressively Grown Their Volume Share of the Chinese Iron Ore Market Over the Last Decade, While Chinese Self-Sufficiency Has Declined Significantly 56

72 Extrapolating the Trend Decline in Self-Sufficiency Forward Suggests the End of the Chinese Iron Ore Industry (and With It the High-Cost Price Support for Iron Ore) in the Region from 2020 to 2025; Clearly, Acceleration in Supply from the Seaborne Market and Declining Steel Growth Rates Would Advance This 56

73 Chinese Growth in Iron Ore (on a Rich-Ore-Equivalent Basis) Peaked in 2007; Since Then, Declining Grades and Rising Costs Have Seen Useable Ore Production Decline 57

74 An Ever-Increasing Amount of Labor Has to Be Allocated to China's Iron Ore Production… 57

75 …Leading to a Prolonged Decline in the Productivity of the Chinese Domestic Industry Since 2007 58

76 An Australian Miner Is 35x as Productive as His Chinese Counterpart; However, Given That Globalization of the Iron Ore Industry Has Enabled China to Outsource Its Iron Ore Mining to Australia, Does That Differential Matter? 58

77 Productivity Uplift of the Iron Ore Imports Into China Has Enabled It to Overtake the U.S. in Terms of Effective Productivity in 2003 59

78 We Compare the Rate at Which Steel Is Being Embedded in China Relative to Peak U.S. Steel Intensity (Y Axis) and the Peak Iron Ore Production (in Man-Hour Terms) to Which China Has Access Relative to That Accessible to the U.S. During Its Industrialization (X Axis); However, Due to Productivity Improvements, the Real Resources Consumed to Embed Steel in the Chinese Economy Are Markedly Lower Than Was the Case During the Industrialization of the U.S. 59

79 U.S. Iron Ore Production Peaked in 1951, With an American Miner Producing 3,150 Tons/Year and Steel Intensity of 33kg/$1,000; Today, China's Steel Intensity Is 63kg/$'000, While the Effective Productivity of Iron Ore Production Is 13,600 Tons/Man-Year 60

80 Bergkamen Coal Washing Plant 61

81 Hückelhoven Coal Washing Plant 61

82 Kaiserstuhl Coking Plant 61

83 Werdohl Leaf Spring Factory 61

84 Assuming Disciplined Capex, We Expect Rio Tinto to Have Room to De-Gear Its Balance Sheet Significantly Over the Next Few Years 66

85 Curtailment of Capex and De-Gearing of the Balance Sheet Would Enable Rio to Raise Its Dividend from US$3.0 Billion in 2012; We See No Reason Why It Should Not Double Over Time 66

86 Rio Tinto's Cost Curve... 67

87 ...Not Dissimilar to Our Own 67

88 Schematically, We Can Represent the Rio Tinto Cost Curve... 68

89 ...And So We Can Calculate the Elasticity of Supply That Rio Tinto Must Believe In 68

90 The Value Destructive Effect of Excessive Volume Growth Can Be Seen Once the Impact of Supply Elasticity Is Considered 69



AND LOVE THE ORE

91 Rio Tinto's Claim That It Does Not See a Price Fall of US$20/t on the Back of Its 70Mtpa Increase in Production Becomes the Key Variable in Deciding Whether or Not the Project Makes Sense; If the Project Induces a Greater Decline, It Will Be Net Value Destructive for the Company 70

92 Rio Tinto Also Believes in a 3% Demand Growth World; This Would Be Sufficient to Limit the Price Impact of the Expansion to ~US$14/t and So Make It Appear Value Accretive (Though Clearly Not by the Degree That It Would Be If the Price Impact of the New Supply Were Neglected in Entirety) 71

93 In Understanding the Value Impact of Consolidation, We Construct Two Simple Industries — One Fragmented... 72

94 ...The Other Consolidated, But in All Other Respects Identical 72

95 For Each Industry, We Simply Look at the Consequences of Each Player Looking to Maximize Apparent Local Value; in This Case, "Apparent" Means Simply Looking at the Economics of Their Own Decision Making; It Is Easy to Show That in a Situation of Potential Oversupply, Local Value Maximization Leads to Global Value Destruction for a Fragmented Industry With No Market Impact Awareness... 73

96 ...And This Global Value Destruction Will Occur Whether or Not Each Individual Player Is Aware of the Impact That Its Decisions Have on the Market Price; the Apparent Gain from Organic Growth in Each Scenario Is an Overwhelmingly Compelling Prize 74

97 In a Consolidated Industry, Where Each Player Acts as If It Was Unaware of the Impact That Its Decisions Have on the Market, There Is Absolutely No Difference in Outcome Versus a Fragmented Industry; Exactly the Same Volume Is Brought to the Market and Exactly the Same Value Destruction Results; Consolidation in Commodity Markets Is Necessary But Not Sufficient to Generate Superior Returns 75

98 The Sufficient Condition for Superior Returns in Consolidated Markets Is an Awareness of the Impact That Volume Decisions Have on Price and the Fact That This Changes the Value of the Entire Portfolio, Not Just the Marginal Ton; Under Such Conditions, the Pursuit of Local Optimality Returns a Global Value Solution in Excess of What Is Optimal 76

99 A Necessary Condition for Superior Price Performance Is That an Industry Be Consolidated, But It Is Not Sufficient; a Necessary and Sufficient Condition Is That There Is a Consolidated Industry Comprising Agents Who Understand the Price Impact of Their Own Decisions 77

100 There Are Really Only Three Players That Matter in Iron Ore (Four at a Push)...This Is a Sufficient Enough Degree of Consolidation That the Pursuit of Local Value Maximization Ought to Embed Capital Discipline in the Industry and Generate Superior Returns 78

101 While the Degree of Possible Oversupply Is Clearly an Important Variable... 78




102 ...Even More Important Is the Elasticity of Supply, and Here Rio Tinto Itself Estimates This to Be About Three; Moreover, It Is the Elasticity of Supply That Inflects Any Commodity Cost Curve and So Generates the Margins That an Industry Enjoys; the Higher the Elasticity, the Higher the Margin and Also the Lower the Degree of Consolidation Required to Generate Superior Returns 79

103 However, This Is Clearly Not What Is Being Discounted in the Stocks 79

104 We Are Beginning to See Greater Clarity from the Miners in Their Thinking on Capital Discipline and Price Preservation...All of Which Ought to Create a Greater Chance of Seeing Growth Restrained and the Industry Re-Rate, Generating Significant Upside for the Miners, Especially Rio Tinto and Vale 80

105 Summary of Historical Commodity Prices 82

106 Except for Some Precious Metals, Current Spot Prices of the Key Commodities for the Miners in Our Coverage Were Substantially Below Their Most Recent Five-Year Mean 82

107 Higher Average Year-to-Date Prices Reflect Higher Commodity Prices in the Beginning of the First Quarter of the Year 83

108 Summary of Commodity Price Scenarios vs. Historical Distribution — Real Terms 84

109 Summary of Commodity Price Scenarios vs. Current Model — Real Terms 84

110 The "Fat Tail" in the Copper Price Distribution Relates to the Period Immediately After the Financial Crisis; We Would Interpret the Shape of the Price Distribution as Telling Us Something About the True Costs of Marginal Production 85

111 The Shift in Pricing Regime from the Benchmark to the Spot Market as Well as the Deterioration in Chinese Domestic Mined Iron Ore Grades Result in a Markedly Different Price Distribution for Iron Ore Than for Copper 85

112 Vale Shows the Greatest Potential Upside and Widest Price Distribution Under Our Scenarios (on the Back of Operational Gearing), While Glencore Shows the Greatest Downside and Least Upside Relative to Our DCF Value (as a Result of Our Bullish View on Copper in the Medium Term) 86

113 Rio Tinto Shows the Least Downside to Our Scenarios and Glencore the Greatest 86

114 As Soon as a More Positive Iron Ore Price Is Introduced, the Impact of Operational Gearing in Vale Is Immediately Apparent 86

115 Our Target Prices Are the Equivalent of the "Bull" Scenario Even Though We Reach These Targets With Higher Short-Term Prices and Lower Long-Term Prices 87

116 It Is Difficult for Us to See Further Upside in Our Target Price for Glencore as It Already Factors in a Very Strong Copper Price 87

117 As This Scenario Analysis Run in Late May Shows, Rio Tinto Offers the Most Attractive Balance Between Risk and Potential Upside 87

118 The Presence of Oil in the BHP Portfolio Adds a Third Critical Variable That Renders It Harder to Unpack the Implied Iron Ore and Copper Price from the Market Price; Oil Remains a Key Diversifier for the Highest Quality Stock in Our Coverage 88



AND LOVE THE ORE

119 While the Market Price of Vale Might Seem to Imply a Higher Iron Ore Price, We Believe That This Is a Function of Us Modeling a Higher Tax Rate for Vale Than the Market 88

120 For Anglo, the Issues of Both Platinum and South African Political Risks Add a Layer of Complexity to the Investment Proposition 88

121 Note That for Glencore Xstrata, We Run the Analysis on Our Combined Pro Forma Glencore Xstrata; Copper Is the Single Greatest Sensitivity for Glencore Xstrata; to See Outperformance in This "Not Yet Institutional Quality" (from a Governance and Reporting Perspective) Stock, It Is Necessary to Believe in a Strong Copper Price 89

122 A Central Question for an Iron Ore Price Forecast Is How to Project China's Steel Intensity Into the Future? Looking at Historical Steel Intensity Alone Is Insufficient 90

123 The Relationship Between Economies' Capital Stock and Level of Output (R-squared = 65%) Provides the Missing Data Point for Forecasting How the Capital Stock to Output Ratio Will Evolve in China 91

124 We Use the U.S. Development Path as a Basis for China's Industrialization Forecast 92

125 The Benefit of Increasing Productivity Over Time Is Easier to See If the Axes in the Earlier Graph Are Inverted; in the U.S., the Real Advances Took Place Only Once the Capital Stock Had Been Built and Urbanization With the Accompanying Labor Force Transformation Had Been Completed 92

126 China Is More Steel Intensive Than the U.S., But Shows Convergence 93

127 For Illustration Only, We Show a Development Path That Would Enable China to Converge to the Current U.S. GDP Levels by 2020; It Would Require a 20% Growth Rate Each Year 94

128 Using Convergence as the "Missing" Data Enables Us to Calculate the Transition Path Between Steel Intensities at Two Different Points in Time and Derive the Relationship Between Overall Economic and Metal Growth Rates 94

129 Since the Time of Our Original Analysis, China's Economy Has Slowed Down Markedly 96

130 In the Space of a Year, the IMF Revised Its Growth Projections Down by 0.5% on Average Over 2014-16) 96

131 Taking the IMF's April 2013 Growth Forecast, We Estimate the Following Convergence to the U.S. Steel Stock Levels for China… 96

132 …Which Would Imply the Following Trajectory for Steel Intensity; the Previous Steel Intensity Trend Line Was Extrapolated in Order to Keep the Long-Term Evolution of China Consistent With the U.S.; Now, We Calculate It Year-by-Year 97

133 We Refine Our Previous Steel Intensity Trend Line… 97

134 …And Generate the Following Steel Production Forecast; It Shows Declines in Crude Steel Growth Seen Post 2020; in Our View, It Is Unavoidable Under All Chinese Growth Scenarios 98

135 Instead of Having to Guess the Relationship Between Metal Growth and Overall Economic Growth, We Can Derive It from the First Principles 98




136 Consensus Expectations for Chinese Steel Production Are Significantly Different from Those Implied by the IMF's GDP Deck 100

137 We Back Out the Scenario for China's GDP Evolution That Is Implied in the Steel Growth Consensus; It Sees the Annual GDP Trend Growth of 7.1 vs. 8.5% Implied in the IMF Deck 100

138 The Steel Intensity That Such a Trajectory Would Imply Is Markedly Lower Than the Scenario Consistent With the IMF GDP Deck 101

139 Instead of a Recovery in the Metal to GDP Multiplier, We Would Continue to See Declines 101

140 Unsurprisingly, Given the Relative Immaturity of the Chinese Economy, Its Steel Growth Is Highly Leveraged (~4 to 1) to the Overall Economic Growth Rate 102

141 Using Our Knowledge of the Structure of the Chinese Domestic Mining Industry, We Can Generate an Exemplar Cost Curve for the Industry as a Whole 103

142 This Enables Us to Understand the Elasticity of Supply 103

143 Combining It With Consensus Steel Demand Expectations Yields Consensus Price Expectations… 104

144 …And Enables Us to Derive Consensus Supply Expectations 104

145 Assuming the Consensus GDP Deck, Two-Thirds of the Supply Expected to Come Onstream Is Not Needed… 104

146 …With Price Destructive Oversupply of ~300Mt by 2020 104

147 We Can Use the Implied Consensus Supply Expectations to Derive Consensus Steel Demand Scenario and Iron Ore Price Line for Any Chinese GDP Deck Assumptions; Steel Intensity Is Geared to China's GDP and the Iron Ore Price Is Geared to Steel Intensity; Risk and Uncertainty Pervades This Sector 106

148 In Order to "Stress Test" the Demand and Price Environment, We Construct Three Scenarios: (1) Slightly More Bullish Than the IMF (i.e., Back to the IMF April 2012 Figures); (2) One That Respects China's Official 7.5% GDP Growth Target; and (3) Significantly Below the Target and the Implied Consensus Growth Rates 108

149 The Resulting Price Paths Illustrate That the Risks for Long-Term Iron Ore Prices Are to the Upside, Barring a Significant "Miss" on the Chinese Growth 108

150 Scenario 1 Uses a Slightly Stronger GDP Deck Than the Current IMF Figures… 109

151 …And Nearly 2% Higher Than Consensus 109

152 To Support This Level of Growth, Steel Intensity Would Need to Rise 109

153 It Would Enable an Accelerated Capital Stock Growth Rate in China 109

154 This Would See Steel Growth Approach 10%... 110

155 ...And a Steel-to-GDP Ratio Well Above 1 110

156 Under This Scenario, Only 18% of New Volume Is Price Destructive… 110

157 …Leading to a Continuation of High Prices Into the Medium Term 110



AND LOVE THE ORE

158 Scenario 2 Assumes the Official Annual GDP Growth Figure of 7.5%... 111

159 …Which Is Still Significantly Ahead of Consensus Expectations 111

160 Although It Can Be Achieved Without an Upward Correction to Steel Intensity… 111

161 …It Still Anticipates an Above-Consensus Demand Growth… 111

162 …With Well-Supported Steel Growth... 112

163 ...And a Declining Steel-to-GDP Ratio 112

164 Nevertheless, 200Mtpa of Supply Is Still Unwarranted from a Value Perspective... 112

165 …Leading to Suppressed Yet Above-Consensus Iron Ore Prices 112

166 Our Bear Case Assumes 6% GDP Growth for China in the Long Term… 113

167 …Which Is ~1% Below the Current Implied Consensus Expectation 113

168 This Would Lead to a Significant Reduction in Steel Intensity... 113

169 ...And Would See Stagnant Production Into the Medium Term 113

170 Consequently, the Industry Would Effectively Go Ex Growth... 114

171 ...And Steel-to-GDP Ratio Would Fall Well Below 1 114

172 In This Scenario, Virtually All Volume Growth Is Price and Value Destructive… 114

173 …Iron Ore Tests a US$65/t Floor 114

174 There Has Been a Significant Change in the Structure of the Chinese Economy Over the Last 10 Years... 116

175 ...With Consumption Losing Out to Investment; Now Investment Accounts for Around Half of All Activity 116

176 Secondary Forms of Economic Activity Have Grown on Average 1.4% Faster Than the Tertiary Forms Over the Course of China's Industrialization (10.7% vs. 9.3%) 117

177 However, the Net Result of All of This Activity Is That China Is on Trend to Catch Up With the U.S. in Terms of the Overall Steel Productivity 117

178 China Currently Has ~35% of the Steel Stock of the U.S.; We Expect It to Rise to ~69% by 2020 118

179 We Use This Relationship Between Steel Stock and Output to Link the Steel Intensity (i.e., Rate of Steel Consumption) Between Different Periods in a Country's Economic Development 118

180 Such a Trajectory of Steel Use Implies FAI Falling from 48% to 38% of GDP, in Effect Achieving the Policy Aim of Rebalancing the Chinese Economy 119

181 However, the Vast Majority of the Investment in China Relates Not to Raw Material Consumption Per Se, But Rather the Capitalization of Labor Associated With That Investment 119

182 Under Our Scenario for Raw Material Consumption (With Constant Prices), the Raw Material Component of FAI Falls from 12.4% (Historical Average) to 8.9% Over the Forecast Period 120

183 Steel Is by Far the Most Important Raw Material Component of FAI; Copper and Aluminum Are Also Significant 120

184 The Proportion of Raw Materials in FAI Has Stayed Very Stable Over Time 121




185 Against This Picture of Raw Material Use, the Increase in the Importance of Investment in China Is Due to the Movement of Labor Into Tertiary Forms of Activity 122

186 Against a Backdrop of Continued De-Agrarianization, a Declining Labor Pool and a Continued Move Into Tertiary Forms of Employment, the Component of the Labor Force in Secondary Forms Must Decline Over Time 122

187 The Trajectory of De-Agrarianization Has China Requiring 15% More Steel Capital Stock Than the U.S. to Effect the Transition to an Economy Dominated by the Service Sector 123

188 This Deceleration in Steel Demand and Change in Composition of the Workforce Imply a Significant Increase in Steel Use Productivity; If Steel Productivity Were Held Flat, There Would Be Significant Upside to Steel Demand 123

189 Against This Demand Environment, We Look at Cost Curves of Iron Ore Landed China; the New Low-Cost Supply Additions Rotate the Cost Curve Downwards; Cost Inflation Translates the Cost Curve Upwards; It Is the Interplay Between Cost Curve Translation and Rotation That Generates Margin and Ultimately Price 124

190 Our Analysis of the Major Miners Informs Our Supply Side Projections 125

191 We Allow for Significant Growth from the Juniors 125

192 FMG Appears to Have the Most Aggressive Growth Plans 125

193 We Then Look at Changes to the Supply and Demand Situation to Imply a Change in the Overall "Balance" of the Iron Ore Industry... 127

194 ...Which Is, Naturally Enough, Different from Consensus 127

195 We Expect a Stronger Demand Environment and a Stronger Supply Side Response Than Consensus 128

196 A Critical Difference Is in the Impact of Expected Continued Cost Escalation on a Non-Static Chinese Iron Ore Industry 128

197 It Is a Combination of the Differences in Supply and Demand Forecasts as Well as Expected Real Cost Escalation in China That Gives Rise to Our Above-Consensus Price Forecast 129

198 Our Iron Ore Price Forecast Is Consistent With the IMF's Current GDP Forecast of 7.8% in 2013 Rising to 8.5% in 2016 and Average Increase in Supply of 88Mtpa 2013-16E 129

199 Roughly 50% of the Upside of Our Forecast vs. Consensus Is Attributable to Differences in Our Cost Escalation Assumptions and 50% Due to S-D Differences 130

200 The First Element of an Iron Ore Price Forecast, in Our View, Is the Calibration of a Consistent View on Steel Intensity and Steel Productivity 131

201 We Calibrate This Against the Overall Composition of Economic Activity and the Labor Force in China 132

202 We Look at How the Cost Curve in China Would Change If There Were to Be No Impact on Price; We Then Calculate the Impact of a Notional "Over Supply" or "Under Supply" on That Cost Curve, Particularly Relative to Consensus Expectations 133

203 This Then Drives Our Price Line 134



AND LOVE THE ORE

204 There Is a Strong Correlation Between the Price of Iron Ore and the Strength of the Australian Dollar, With a High Iron Ore Price Equating to a Strong Local Currency 136

205 Likewise, the Relationship Exists (Albeit Less Pronounced) for the Brazilian Real 136

206 Unfortunately for Anglo American, for the South African Rand, the Relationship Is Not in Evidence 137

207 Mining Costs and Revenue Exhibit a High Degree of Correlation 137

208 Currency Appreciation Drives a Significant Part of the Cost-to-Price Relationship… 137

209 ...Which Is Particularly True for the Australian Producers Rio Tinto and BHP Billiton 137

210 Rio Tinto Is the Lowest Cost Iron Ore Producer in Our Coverage; Anglo, as the Highest, Experiences Cash Opex That Is 50% Higher Than Rio's 139

211 While Mining Cost Escalation Is a Common Worry, It Appears That the Miners Have Started to Get It Back Under Control 139

212 Rio Tinto, in Particular, Is Leading the Way 139

213 BHP Has Gone Through a Similar Experience With Both the Cost Relationship to Iron Ore... 140

214 ...And Over Time, Showing a Marked Deceleration 140

215 For Vale, the Evidence of Cost Control Is Less Convincing... 140

216 …As Can Be Seen in the Continued Cost Growth Over Time 140

217 For Anglo, the Situation Seems to Lie Somewhere Between Vale and the Australians… 141

218 …With Some Evidence of Cost Containment 141

219 An Internet Search for the Miners and "Cost Controls" Results in an Ad for a Cost Control Position as at Least One of the Top 10 Hits (as of April 30, 2013); Only Anglo Has All of Its First Five Hits as Job Openings 141

220 Given the Historical Relationship Between the Iron Ore Price and Producer Currencies, a Return to US$80/t Iron Ore Would See Producer Currencies Weaken, Lowering US$ Costs and Improving Margin for the Miners 142

221 The Sensitivity to Such a Margin Impact Gives, on Average, ~6% Additional Upside to the Australian and Brazilian Miners But Is Value Neutral for Anglo American 142

222 Rusting Granite in Peterborough, New Hampshire; During Pre-Industrial and Early Industrial Periods, Local Blacksmiths Relied Upon "Bog Iron" and "Mountain Iron" (Like the Below); While the Highest-Quality Deposit Reportedly Contained Ore of 56-63%, the Grade Was Often Much Lower and Was Quickly Depleted in Concentrations of Useful Economic Value 144

223 Where All the EBITDA Began — A Banded Iron Formation (BIF) Showing Characteristic Layering Patterns of Iron Ore and Gangue Material 145

224 Not All Iron Ores Are Created Equal... 145

225 …And There Is a Significant "Tail" of Non-Iron Material in Iron Ores, All of Which Have a Value Implication 146




226 While Australia and Brazil Are the Most Important Suppliers Into China, They Account for Only ~70% of the Total Non-Chinese Portion of Supply 147

227 The Overall Market Shares of Brazil and Australia Have Remained Roughly Constant Over Time; Perhaps This Indicates an Understanding from Incumbents That Competing on Volume in Commodity Markets Is Value Destructive 147

228 There Is a Considerable Variation in the Observed Grade of Iron Ore Landed Into China — It Is Still Far from Being a Homogeneous Product 148

229 An Analysis of 208 Pricing Points Shows a Significant Range of Variation Around the Benchmark 62% CIF Price Point 148

230 Unsurprisingly, Fe Grade Explains Most of the Variation in Price 149

231 Nevertheless, Even After Adjusting for Fe Grade, Poor Ores Suffer a Significant Discount 149

232 This Is True Across Australian Iron Ores... 149

233 ...As Well as Those from India 149

234 There Is a US$46/t Differential Across Ores from Different Supply Locations 150

235 Across the "Major" Space, There Is a US$25/t Differential 150

236 Assuming That 50% of the Quality-Disadvantaged Material Moves Out of the Market as Price Begins to Fall, It Can Be Expected to Provide ~US$5/t of Offsetting Upward Price Support, Adding, on Average, $1.75/Share Across Our Coverage (as of June 30, 2013 Prices) 151

237 Following the Same Three-Phase Trajectory for the Evolution of the Steel Consumption Rate Within All Industrialized and Industrializing Countries, China's Steel Intensity Has Passed from "Development" Through "Peak" and Is Now in "Decline"; the Steady State Rate of Consumption, Once the Process Is Completed, Will Be Determined by Whether the Economy, Once Mature, Adopts a More Manufacturing- or Services-Oriented Focus 155

238 In 2000, China Represented a Clear Outlier from the Normal Pattern: the World's Largest Country Was Set to Age Rapidly and Yet Was Still at a Pre-Industrial Level of Output; by 2020, China's Working-Age Population Will Have Peaked; Over the Next Generation, 260 Million People Will Leave the Chinese Labor Force, Leading to a 90% Increase in the Old-Age Dependency Ratio 156

239 Over the Last 10 Years, the Domestic Steel Price in China Has Been Remarkably Constant... 157

240 ...Whereas the Price of Raw Materials Has Accelerated Sharply; It Is True for All — Iron Ore... 157

241 ...Key Fuel and Reductants Used in Steel Making — HCC... 158

242 ...And Its Replacement PCI 158

243 The Rise in Iron Ore Price Has Done the Most to Increase the Costs of Chinese Steel Production, Reflecting a Transfer of Value from the Chinese Industrial Sector to the Western-Owned Miners 158

244 Raw Materials Fed Into Steel Making Used to Cost Less Than 100US$/t; It Is Now Well Over 300US$/t 159



AND LOVE THE ORE

245 Currently, the Raw Material Cost of Steel Making Accounts for Nearly Three Quarters of the Overall Costs in China 159

246 There Has Also Been a Slower, But Nonetheless Pronounced Rise in the Non-Raw Materials Costs of Steel Making in China 159

247 Unsurprisingly, Margins in the Chinese Steel-Making Industry Have Declined on the Back of Higher Raw Material Prices 160

248 While, at First Glance, the Growth in Chinese Steel Production Appears Monotonic... 160

249 ...This Belies a Marked Slowdown in Chinese Steel Production Rates from ~20% at the Start of the Last Decade to Closer to 6% Today, Concurrent With the Declining Steel-Making Margins 161

250 Growth and Profitability Declined Concurrently 161

251 We Believe That a Significant Part of the Slowdown in Chinese Steel Growth Resulted from High Commodity Prices; the West Experienced a Step Change in Steel Demand Growth Only After Urbanization Was Complete; However, This Is Far from Complete in China, Combined With the Imminent Demographic Shift Enables Us to Conclude That in the Presence of Falling Iron Price, Steel Demand Would Rise 162

252 A Complicating Factor in the Analysis Steel Elasticity Is the Effect of Chinese Monetary Stimulus in 2009 163

253 There Is a Strong Relationship Between Lagged Steel Growth and the Growth in the Supply of Money... 163

254 ...We Strip This Effect Out of the Analysis of Steel Demand Growth... 163

255 ...The Effect of This Would Be to Lower the Level of Steel Production for Any Given Profitability... 164

256 ...Which Increases the Strength of the Relationship 164

257 It Is Inconsistent to Simultaneously Believe That Iron Ore Prices Would Fall as a Consequence of Low Chinese Steel Growth and That China Is Undergoing a Period of Ongoing Capital Stock Accumulation; Consequently, We Believe That Falling Iron Ore Prices Would Lead to Increased Steel Growth of an Additional 3% Impact on Trend Steel Growth Rates, Even Given a Declining Steel Price 165

258 The Impact of the Additional Steel Demand Helping Offset the Impact of "Oversupply" from New Projects Could Support a Higher Price Level Than Would Otherwise Be Achieved 166

259 The Impact of Incorporating This Effect Could See Iron Ore Prices Stabilize Some 15US$/t Higher Than Would Be the Case If the Elasticity of Demand Is Ignored 166

260 Summary of Our Coverage 169

261 The Relationship Between the Miners and Their Underlying Commodity Exposure Is Incredibly Strong... 170

262 ...And Explains the Majority of the Historical Performance 170

263 BHP Is Currently Slightly Overvalued Relative to Its Historical Best Fit, an Unsurprising Development in the Current Tremulous Market Environment 171

264 We See Considerable Upside in the Share Price 171

265 The Relationship Between the Miners and Their Underlying Commodity Exposure Is Incredibly Strong… 172

266 …And Explains the Majority of the Historical Performance 172




267 Rio Is Currently Slightly Overvalued Relative to Its Historical Best Fit, But Less Than 5%, and Hence Less Than High-Quality BHP 173




271 Anglo Is Currently Slightly Undervalued Relative to Its Historical Best Fit, But We Note the Regression Has Shown Instability and a Decision to Invest in the Stock Is Contingent Upon a Belief in the Fundamental Restructuring Story as Well as Commodity Prices 175




275 Vale Is Currently Slightly Undervalued Relative to Its Historical Best Fit, But as With Anglo, We Caution That the Regression Has Been Unstable and That There Are Risks Not Captured by It (Specifically the Significant Influence of the Brazilian Government) 177


277 For Glencore Xstrata, the Historical Relationship Is Based on a "Synthetic" Stock Using the Xstrata Price History and Recent Relationship to Glencore 179

278 And in Glencore Xstrata (Though for Glencore Xstrata the Historical Relationship Is Based on a "Synthetic" Stock Using the Xstrata Price History and Recent Relationship to Glencore) 179

279 Glencore Xstrata Seems Fairly Valued Relative to the Historical Best Fit of This Synthetic Stock 180


281 BHP DCF Value 182

282 Rio Tinto DCF Value 182

283 Anglo American DCF Value 183

284 Vale DCF Value 183

285 Glencore Xstrata DCF Value (With and Without Synergies) 184

286 Iron Ore Represents, by Far, the Most Significant Contribution to the Cash Generation of the Miners… 187

287 …And Has Been, Along With Copper, the Strongest Performer of the Major Commodities 187

288 Vale Represents Essentially an Iron Ore Pure Play in Both the Near and Longer Term, While We See Rio Diversifying in the Medium Term; BHP and Anglo American Have Exposure to a Well-Diversified Basket of Commodities; Glencore Xstrata Is the Only Company in Our Coverage Without Any Direct Production Exposure to Iron Ore* 188

289 Vale Is the Most Heavily Exposed to Iron Ore of Our Coverage Group (a Trend We Do Not See Changing)… 189

290 …Followed by Rio Tinto, Which Is Augmenting Its Portfolio With Some of the World's Best Copper Projects (the Metal on Which We Are the Most Bullish) 189

291 Historical Evolution of Rio Tinto's Revenue by Business Unit 189

292 Historical Evolution of Rio Tinto's EBITDA by Business Unit 190



AND LOVE THE ORE

293 Historical Evolution of Vale's Revenue by Business Unit 190

294 Historical Evolution of Vale's EBITDA by Business Unit 191

295 Our Forecasts for Rio Tinto's Revenues… 192

296 …And EBITDA Through 2018 See Considerable Growth 192

297 Similarly, Our Forecasts for Vale's Revenues… 192

298 …And EBITDA Also Sees Considerable Growth 192

299 The Diversity in Exposure Makes BHP One of the Lowest-Risk Miners from a Commodity Perspective… 193

300 …Complemented by Low Political Risk and a Diverse Range of Operating Geographies 193

301 Historical Evolution of BHP's Revenue by Business Unit 194

302 Historical Evolution of BHP's EBITDA by Business Unit 194

303 BHP's Petroleum Exposure Is Its Great Diversifier (Generating 69% EBITDA Margins in 2012)… 195

304 …While Anglo American's Great Diversifier, Platinum, Came in at Margins of Just 11% 195

305 We Also See Revenue and EBITDA Growth… 195

306 …For BHP Billiton 195

307 Anglo and Glencore Xstrata Represent the Tier 2 of Mining Companies Behind the "Big Three" of Vale, Rio and BHP 197

308 The Differentiating Feature of the "Big Three," Both in Terms of Size and Margin, Is the Presence of Significant Iron Ore Exposure 197

309 Anglo Is the Most Diversified and the Lowest-Risk Miner from a Commodity Exposure Perspective 197

310 The Relative Lack of "Safe" Australian Exposure Has Always Been a Risk Compared to the Other Miners 197

311 Historical Evolution of Anglo American's Revenue by Business Unit 198

312 Historical Evolution of Anglo American's EBITDA by Business Unit 198

313 An Increasing Proportion of Anglo Platinum's Refined Production Now Comes from Non-Mined Sources… 199

314 …The Majority Is Purchased from Related Joint Venture Partners and Associates 199

315 Our Forecast for Anglo American Sees Revenue Growth… 199

316 …As Well as EBITDA Growth as Part of the Turnaround Story 199

317 Glencore Xstrata's High-Revenue, Low-Margin Trading Business Renders Its Group Profile Very Different from the Pure Miners in Our Coverage… 201

318 …Though Excluding the Marketing Division Reveals a Profile That Is Similar to the Rest of the Miners 201

319 The Impact of the Marketing Division Can Be Seen in Total Company Margins… 201

320 …While the Margins for Only Industrial Activities More Closely Resemble the Pure Miners in Our Coverage 201

321 Historical Pro Forma Evolution of Glencore Xstrata's Industrial Activities EBITDA 202

322 Glencore Xstrata's Industrial Activities Are Predominantly a Copper, Coal and Zinc Play… 202




323 …As Can Also Be Seen When Looking at the Company in Toto (Inclusive of Marketing) 202

324 Vale Shows the Greatest Potential Upside and Widest Price Distribution Under Our Scenarios (on the Back of Operational Gearing), While Glencore Shows the Greatest Downside and Least Upside Relative to Our Current Price Target (as a Result of Our Bullish View on Copper in the Medium Term) 205

325 Rio Tinto Shows the Least Downside to Our Scenarios and Glencore the Greatest 206

326 As Soon as a More Positive Iron Ore Price Is Introduced, the Impact of Operational Gearing in Vale Is Immediately Apparent 206

327 Our Target Prices Are the Equivalent of the "Bull" Scenario Even Though We Reach These Targets With Higher Short-Term Prices and Lower Long-Term Prices 206

328 It Is Difficult for Us to See Further Upside in Our Target Price for Glencore as It Already Factors in a Very Strong Copper Price 206

329 Rio Tinto Scenario Summary vs. Consensus 207

330 BHP Billiton Scenario Summary vs. Consensus 207

331 Anglo American Scenario Summary vs. Consensus 207

332 Glencore Xstrata Pro Forma Scenario Summary vs. Consensus 208

333 Vale Scenario Summary vs. Consensus 208

334 Both the "Grizzly"... 209

335 ...And "Bear" Scenarios Show Revenue Below Consensus Expectations 209

336 Over the Next Two Years What We Call the "Bull" Scenario Is Closest to Consensus Expectation 209

337 While Under the Strongest Price Scenario, There Is Upside to Consensus 209

338 The Same Feature Is Seen at the EBITDA Line... 210

339 ...The Lack of Coverage of Glencore Makes Comparison More Difficult 210

340 Again, the "Bull" Scenario Appears Closest to Consensus 210

341 Considerable Upside Under Stronger Commodity Scenarios Emphasizes the Degree of Operationally Geared Exposure the Miners Offer to Commodity Prices 210

342 Rio Tinto Scenario Analysis Summary 211

343 Rio Tinto EV by Division 211

344 Rio Tinto — Current Model — Pro Forma Financials 212

345 Rio Tinto — Current Model — Valuation 212

346 Rio Tinto — Grizzly — Pro Forma Financials 213

347 Rio Tinto — Grizzly — Valuation 213

348 Rio Tinto — Bear — Pro Forma Financials 214

349 Rio Tinto — Bear — Valuation 214

350 Rio Tinto — Bull — Pro Forma Financials 215

351 Rio Tinto — Bull — Valuation 215

352 Rio Tinto — De Niro — Pro Forma Financials 216

353 Rio Tinto — De Niro — Valuation 216

354 BHP Billiton Scenario Analysis Summary 217

355 BHP Billiton EV by Division 217



AND LOVE THE ORE

356 BHP Billiton — Current Model — Pro Forma Financials 218

357 BHP Billiton — Current Model — Valuation 218

358 BHP Billiton — Grizzly — Pro Forma Financials 219

359 BHP Billiton — Grizzly — Valuation 219

360 BHP Billiton — Bear — Pro Forma Financials 220

361 BHP Billiton — Bear — Valuation 221

362 BHP Billiton — Bull — Pro Forma Financials 222

363 BHP Billiton — Bull — Valuation 222

364 BHP Billiton — De Niro — Pro Forma Financials 223

365 BHP Billiton — De Niro — Valuation 223

366 Anglo American Scenario Analysis Summary 224

367 Anglo American EV by Division 224

368 Anglo American — Current Model — Pro Forma Financials 225

369 Anglo American — Current Model — Valuation 225

370 Anglo American — Grizzly — Pro Forma Financials 226

371 Anglo American — Grizzly — Valuation 226

372 Anglo American — Bear — Pro Forma Financials 227

373 Anglo American — Bear — Valuation 227

374 Anglo American — Bull — Pro Forma Financials 228

375 Anglo American — Bull — Valuation 228

376 Anglo American — De Niro — Pro Forma Financials 229

377 Anglo American — De Niro — Valuation 229

378 Glencore Xstrata Pro Forma Scenario Analysis Summary 230

379 Glencore Xstrata Pro Forma EV by Division 230

380 Glencore Xstrata Pro Forma — Current Model — Pro Forma Financials 231

381 Glencore Xstrata Pro Forma — Current Model — Valuation 232

382 Glencore Xstrata Pro Forma — Grizzly — Pro Forma Financials 233

383 Glencore Xstrata — Grizzly — Valuation 234

384 Glencore Xstrata Pro Forma — Bear — Pro Forma Financials 235

385 Glencore Xstrata Pro Forma — Bear — Valuation 236

386 Glencore Xstrata — Bull — Pro Forma Financials 237

387 Glencore Xstrata Pro Forma — Bull — Valuation 238

388 Glencore Xstrata — De Niro — Pro Forma Financials 239

389 Glencore Xstrata — De Niro — Valuation 240

390 Vale Scenario Analysis Summary 241

391 Vale Pro Forma EV by Division 241

392 Vale — Current Model — Pro Forma Financials 242

393 Vale — Current Model — Valuation 242

394 Vale — Grizzly — Pro Forma Financials 243

395 Vale — Grizzly — Valuation 243

396 Vale — Bear — Pro Forma Financials 244

397 Vale — Bear — Valuation 244

398 Vale — Bull — Pro Forma Financials 245

399 Vale — Bull — Valuation 245

400 Vale — De Niro — Pro Forma Financials 246




401 Vale — De Niro — Valuation 246

402 Long-Term Real Commodity Price Scenarios 247

403 Nickel Spot Price Distribution 248

404 Zinc Spot Price Distribution 248

405 Aluminum Spot Price Distribution 249

406 Platinum Spot Price Distribution 249

407 Palladium Spot Price Distribution 250

408 Gold Spot Price Distribution 250

409 Silver Spot Price Distribution 251

410 Thermal Coal Spot Price Distribution 251

411 Molybdenum Spot Price Distribution 252

412 Uranium Spot Price Distribution 252

413 Consensus Scenario 254

414 IMF Scenario 255

415 Back to the Future 256

416 Ending With a Whimper 257

417 It's All Over Now Baby Blue 258

418 Rio Tinto Income Statement 259

419 Rio Tinto Balance Sheet 260

420 Rio Tinto Cash Flow Statement 261

421 BHP Income Statement 262

422 BHP Balance Sheet 263

423 BHP Cash Flow Statement 264

424 Vale Income Statement 265

425 Vale Balance Sheet 266

426 Vale Cash Flow Statement 267

427 Anglo American Income Statement 268

428 Anglo American Balance Sheet 269

429 Anglo American Cash Flow Statement 270

430 Glencore Xstrata Pro Forma Income Statement (Including Synergies) 271

431 Glencore Xstrata Pro Forma Balance Sheet (Including Synergies) 272

432 Glencore Xstrata Pro Forma Cash Flow Statement (Including Synergies) 273

433 Glencore Xstrata Pro Forma Income Statement (Excluding Synergies) 274

434 Glencore Xstrata Pro Forma Balance Sheet (Excluding Synergies) 275

435 Glencore Xstrata Pro Forma Cash Flow (Excluding Synergies) 276



AND LOVE THE ORE

Disclosure Appendix

VALUATION METHODOLOGY

Mining companies are operationally and financially geared to their underlying commodity exposure, so we provide two valuation metrics (see the "Valuation and Risks" chapter of this Blackbook for more details):

• ~80% of weekly mining equity price moves can be explained by underlying commodity price moves, so we use a regression-based trading model and our commodity price forecasts to help determine our 12-month price targets. If the regression remains stable or deviations appear temporary, the model determines the target price. If we believe a deviation is signaling a fundamental change, we will adjust our target price for this fundamental shift and disclose the manner and magnitude of the adjustment made. At present, no adjustments have been made.

• We additionally provide a supplementary DCF-based valuation constructed in nominal local currency terms out to 2030 over which explicit commodity price and exchange rate forecasts apply. The nominal local currency cash flows are de-escalated into real U.S. dollar cash flows and discounted at the company-specific WACC. A country risk premium reflecting the geographic origin of the cash flows is added to the underlying WACC to reflect cash flow items (i.e., expropriation) that cannot be explicitly modeled in the cash flow. All reserves are considered exploited by the model. In addition, 50% of the incremental resources (i.e., 50% of the residual resources, excluding those that have already been converted to reserves) of the company are modeled. Where residual life of mine (LOM) may be inferred for operations beyond the 2030 time horizon, a terminal value is applied for the remaining years of potentially exploitable material. We forecast our models in reporting currency (USD), convert to listing currency (GBP or Real), and round final DCF values in 25p/cent increments.

RISKS

The four most significant risks facing the major mining houses are: 1) lack of capital discipline (specifically displacement of high-cost Chinese marginal producers by low-cost Western production), 2) operating cost inflation (U.S. dollar denominated unit costs in all the major mining houses have seen double-digit growth rates over the last 10 years, roughly half of which are macro related and the other half are real local currency), 3) a sustained downturn in the Chinese economy (the largest consumer of global resources), and 4) resource nationalism (ranging from increased share of rent extraction to outright asset confiscation). For more details on both sector risks and company-specific risks, see the risk section in the "Valuation and Risks" chapter.

SRO REQUIRED DISCLOSURES

References to "Bernstein" relate to Sanford C. Bernstein & Co., LLC, Sanford C. Bernstein Limited, Sanford C. Bernstein (Hong Kong) Limited, and Sanford C. Bernstein (business registration number 53193989L), a unit of AllianceBernstein (Singapore) Ltd. which is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C, collectively.

Bernstein analysts are compensated based on aggregate contributions to the research franchise as measured by account penetration, productivity and proactivity of investment ideas. No analysts are compensated based on performance in, or contributions to, generating investment banking revenues.

Bernstein rates stocks based on forecasts of relative performance for the next 6-12 months versus the S&P 500 for stocks listed on the U.S. and Canadian exchanges, versus the MSCI Pan Europe Index for stocks listed on the European exchanges (except for Russian companies), versus the MSCI Emerging Markets Index for Russian companies and stocks listed on emerging markets exchanges outside of the Asia Pacific region, and versus the MSCI Asia Pacific ex-Japan Index for stocks listed on the Asian (ex-Japan) exchanges - unless otherwise specified. We have three categories of ratings:

Outperform: Stock will outpace the market index by more than 15 pp in the year ahead.

Market-Perform: Stock will perform in line with the market index to within +/-15 pp in the year ahead.

Underperform: Stock will trail the performance of the market index by more than 15 pp in the year ahead.

Not Rated: The stock Rating, Target Price and estimates (if any) have been suspended temporarily.

As of 07/11/2013, Bernstein's ratings were distributed as follows: Outperform - 39.6% (0.9% banking clients) ; Market-Perform - 47.2% (0.0% banking clients); Underperform - 13.2% (0.0% banking clients); Not Rated - 0.0% (0.0% banking clients). The numbers in parentheses represent the percentage of companies in each category to whom Bernstein provided investment banking services within the last twelve (12) months.

Accounts over which Bernstein and/or their affiliates exercise investment discretion own more than 1% of the outstanding common stock of the following companies RIO.LN / Rio Tinto PLC, BLT.LN / BHP Billiton PLC.

This research publication covers six or more companies. For price chart disclosures, please visit www.bernsteinresearch.com, you can also write to either: Sanford C. Bernstein & Co. LLC, Director of Compliance, 1345 Avenue of the Americas, New York, N.Y. 10105 or Sanford C. Bernstein Limited, Director of Compliance, 50 Berkeley Street, London W1J 8SB, United Kingdom; or Sanford C. Bernstein (Hong Kong) Limited, Director of Compliance, Suites 3206-11, 32/F, One International Finance Centre, 1 Harbour View Street, Central, Hong Kong, or Sanford C. Bernstein (business registration number 53193989L) , a unit of AllianceBernstein (Singapore) Ltd. which is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C, Director of Compliance, 30 Cecil Street, #28-08 Prudential Tower, Singapore 049712.




12-Month Rating History as of 07/14/2013

Ticker Rating Changes

AAL.LN O (IC) 09/05/12 O (DC) 07/29/11

BBL O (IC) 09/26/12 O (DC) 07/29/11

BHP O (IC) 09/05/12

BHP.AU O (IC) 09/26/12

BLT.LN O (IC) 09/05/12 O (DC) 07/29/11

GLEN.LN O (RC) 02/13/13 M (IC) 09/05/12

RIO O (IC) 09/05/12 O (DC) 07/29/11

RIO.LN O (IC) 09/05/12 O (DC) 07/29/11

VALE O (RC) 06/07/13 U (IC) 09/05/12

VALE3.BZ O (RC) 06/07/13 U (IC) 09/05/12

Rating Guide: O - Outperform, M - Market-Perform, U - Underperform, N - Not Rated

Rating Actions: IC - Initiated Coverage, DC - Dropped Coverage, RC - Rating Change

OTHER DISCLOSURES

A price movement of a security which may be temporary will not necessarily trigger a recommendation change. Bernstein will advise as and when coverage of securities commences and ceases. Bernstein has no policy or standard as to the frequency of any updates or changes to its coverage policies. Although the definition and application of these methods are based on generally accepted industry practices and models, please note that there is a range of reasonable variations within these models. The application of models typically depends on forecasts of a range of economic variables, which may include, but not limited to, interest rates, exchange rates, earnings, cash flows and risk factors that are subject to uncertainty and also may change over time. Any valuation is dependent upon the subjective opinion of the analysts carrying out this valuation.

This document may not be passed on to any person in the United Kingdom (i) who is a retail client (ii) unless that person or entity qualifies as an authorised person or exempt person within the meaning of section 19 of the UK Financial Services and Markets Act 2000 (the "Act"), or qualifies as a person to whom the financial promotion restriction imposed by the Act does not apply by virtue of the Financial Services and Markets Act 2000 (Financial Promotion) Order 2005, or is a person classified as an "professional client" for the purposes of the Conduct of Business Rules of the Financial Conduct Authority.

To our readers in the United States: Sanford C. Bernstein & Co., LLC is distributing this publication in the United States and accepts responsibility for its contents. Any U.S. person receiving this publication and wishing to effect securities transactions in any security discussed herein should do so only through Sanford C. Bernstein & Co., LLC.

To our readers in the United Kingdom: This publication has been issued or approved for issue in the United Kingdom by Sanford C. Bernstein Limited, authorised and regulated by the Financial Conduct Authority and located at 50 Berkeley Street, London W1J 8SB, +44 (0)20-7170-5000.

To our readers in member states of the EEA: This publication is being distributed in the EEA by Sanford C. Bernstein Limited, which is authorised and regulated in the United Kingdom by the Financial Conduct Authority and holds a passport under the Markets in Financial Instruments Directive.

To our readers in Hong Kong: This publication is being distributed in Hong Kong by Sanford C. Bernstein (Hong Kong) Limited which is licensed and regulated by the Hong Kong Securities and Futures Commission (Central Entity No. AXC846). This publication is solely for professional investors only, as defined in the Securities and Futures Ordinance (Cap. 571).

To our readers in Singapore: This publication is being distributed in Singapore by Sanford C. Bernstein, a unit of AllianceBernstein (Singapore) Ltd., only to accredited investors or institutional investors, as defined in the Securities and Futures Act (Chapter 289). Recipients in Singapore should contact AllianceBernstein (Singapore) Ltd. in respect of matters arising from, or in connection with, this publication. AllianceBernstein (Singapore) Ltd. is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C. It is regulated by the Monetary Authority of Singapore and located at 30 Cecil Street, #28-08 Prudential Tower, Singapore 049712, +65-62304600. The business name "Sanford C. Bernstein" is registered under business registration number 53193989L.

To our readers in Australia: Sanford C. Bernstein & Co., LLC, Sanford C. Bernstein Limited and Sanford C. Bernstein (Hong Kong) Limited are exempt from the requirement to hold an Australian financial services licence under the Corporations Act 2001 in respect of the provision of the following financial services to wholesale clients:

providing financial product advice;

dealing in a financial product;

making a market for a financial product; and

providing a custodial or depository service.

Sanford C. Bernstein & Co., LLC., Sanford C. Bernstein Limited, Sanford C. Bernstein (Hong Kong) Limited and AllianceBernstein (Singapore) Ltd. are regulated by, respectively, the Securities and Exchange Commission under U.S. laws, by the Financial Conduct Authority under U.K. laws, by the Hong Kong Securities and Futures Commission under Hong Kong laws, and by the Monetary Authority of Singapore under Singapore laws, all of which differ from Australian laws.



AND LOVE THE ORE

One or more of the officers, directors, or employees of Sanford C. Bernstein & Co., LLC, Sanford C. Bernstein Limited, Sanford C. Bernstein (Hong Kong) Limited, Sanford C. Bernstein (business registration number 53193989L) , a unit of AllianceBernstein (Singapore) Ltd. which is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C, and/or their affiliates may at any time hold, increase or decrease positions in securities of any company mentioned herein.

Bernstein or its affiliates may provide investment management or other services to the pension or profit sharing plans, or employees of any company mentioned herein, and may give advice to others as to investments in such companies. These entities may effect transactions that are similar to or different from those recommended herein.

Bernstein Research Publications are disseminated to our customers through posting on the firm's password protected website, www.bernsteinresearch.com. Additionally, Bernstein Research Publications are available through email, postal mail and commercial research portals. If you wish to alter your current distribution method, please contact your salesperson for details.

Bernstein and/or its affiliates do and seek to do business with companies covered in its research publications. As a result, investors should be aware that Bernstein and/or its affiliates may have a conflict of interest that could affect the objectivity of this publication. Investors should consider this publication as only a single factor in making their investment decisions.

This publication has been published and distributed in accordance with Bernstein's policy for management of conflicts of interest in investment research, a copy of which is available from Sanford C. Bernstein & Co., LLC, Director of Compliance, 1345 Avenue of the Americas, New York, N.Y. 10105, Sanford C. Bernstein Limited, Director of Compliance, 50 Berkeley Street, London W1J 8SB, United Kingdom, or Sanford C. Bernstein (Hong Kong) Limited, Director of Compliance, Suites 3206-11, 32/F, One International Finance Centre, 1 Harbour View Street, Central, Hong Kong, or Sanford C. Bernstein (business registration number 53193989L) , a unit of AllianceBernstein (Singapore) Ltd. which is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C, Director of Compliance, 30 Cecil Street, #28-08 Prudential Tower, Singapore 049712. Additional disclosures and information regarding Bernstein's business are available on our website www.bernsteinresearch.com.

CERTIFICATIONS

I/(we), Paul Gait, Senior Analyst(s)/Analyst(s), certify that all of the views expressed in this publication accurately reflect my/(our) personal views about any and all of the subject securities or issuers and that no part of my/(our) compensation was, is, or will be, directly or indirectly, related to the specific recommendations or views in this publication.

Approved By: NK

Copyright 2013, Sanford C. Bernstein & Co., LLC, Sanford C. Bernstein Limited, Sanford C. Bernstein (Hong Kong) Limited, and AllianceBernstein (Singapore) Ltd., subsidiaries of AllianceBernstein L.P. ~1345 Avenue of the Americas ~ NY, NY 10105 ~212/756-4400. All rights reserved.

This publication is not directed to, or intended for distribution to or use by, any person or entity who is a citizen or resident of, or located in any locality, state, country or other jurisdiction where such distribution, publication, availability or use would be contrary to law or regulation or which would subject Bernstein or any of their subsidiaries or affiliates to any registration or licensing requirement within such jurisdiction. This publication is based upon public sources we believe to be reliable, but no representation is made by us that the publication is accurate or complete. We do not undertake to advise you of any change in the reported information or in the opinions herein. This publication was prepared and issued by Bernstein for distribution to eligible counterparties or professional clients. This publication is not an offer to buy or sell any security, and it does not constitute investment, legal or tax advice. The investments referred to herein may not be suitable for you. Investors must make their own investment decisions in consultation with their professional advisors in light of their specific circumstances. The value of investments may fluctuate, and investments that are denominated in foreign currencies may fluctuate in value as a result of exposure to exchange rate movements. Information about past performance of an investment is not necessarily a guide to, indicator of, or assurance of, future performance.


EUROPEAN METALS & MINING: A STRANGE LOVE – HOW I LEARNED TO STOP WORRYING AND LOVE THE ORE


SO

N LA

PO

RT

E at P

ER

RY

CA

PIT

AL on 16-Jul-2013

iron ore primer

Documents