Raphael Trotignon - PhD Defense Under the supervision of Professor Christian de Perthuis

THE EUROPEAN UNION CO2 EMISSION TRADING SCHEME:

FROM EX ANTE AND EX POST ANALYSIS TO THE PROJECTION IN 2020

October 17th, 2012 – Paris Dauphine University (D520)

Climate Economics Chair - raphael.trotignon@chaireeconomieduclimat.org

IN SEARCH OF THE CARBON PRICE

mailto:raphael.trotignon@chaireeconomieduclimat.org

2

Introduction

• This thesis is the result of five years studying the European Emission Trading Scheme, the most complete experience of carbon pricing to date • An empirical analysis based on ex ante/ex post comparison • Supplemented by the creation of a simulation model, an original approach • In a context full of uncertainties, nobody knows the “right” carbon price: it is necessary to combine the three core flexibility mechanisms associated to the EU ETS to understand the market development (Chapter I) • Analysis build by progressively integrating in the model: baseline emissions and abatement (Chapter II), the use of offsets (Chapter III), and banking and borrowing (Chapter IV) • Eventually, shift from an ex post use to a prospective use of the model to 2020 (Chapter V)

3

I. Ex ante vs. ex post

Beginning of 2005 Beginning of 2008 Beginning of 2012

0

500

1 000

1 500

2 000

2 500

3 000

3 500

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

20

17

20

18

20

19

20

20Mill

ion

s

Total cap (free+auction+offsets)

Cap (free+auction)

Baseline as in 2005

- 2 GtCO2

- 350 MtCO2

0

500

1 000

1 500

2 000

2 500

3 000

3 500

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

20

17

20

18

20

19

20

20

Mill

ion

sTotal cap (free+auction+offsets)

Cap (free+auction)

Verified Emissions

Baseline as in 2008

- 500 MtCO2 - 5 GtCO2

0

500

1 000

1 500

2 000

2 500

3 000

3 500

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

20

17

20

18

20

19

20

20

Mill

ion

s

Total cap (free+auction+offsets)

Cap (free+auction)

Verified Emissions

Baseline as in 2010

- 850 MtCO2+ 1,2 GtCO2

+ 225 MtCO2

• Difference between initial expectations and actual outcome: – Past emission reductions (carbon price and economic crisis) – Decentralized use of offsets, supplemental to the cap – Interactions with other policies – Weakened anticipations

4

I. The observed price

Source: Climate Economics Chair

0

5

10

15

20

25

30

35

40

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

OTC Point Carbon BNS EUA05-07 ICE ECX EUA DEC12

Expectations in line with early analysts forecasts Illiquid market

EU ETS Directive adoption

Decisions on NAP1

Cold winter

Phase 1 surplus not transferable to Phase 2

Release of 2005 verified emissions

Discussions on 2020 targetsNAP 2 decisions

Financial and economic crisis

New equilibirum in a context of slow recovery

Roadmap 2050Fukushima Energy

efficiency

Debt crisis and worsening growth outlookSet aside debate

Climate Energy Package

Phase 1 spot price December 2012 Future price

• The price is very different from initial expectations • A “classic” for cap-and-trade programs (SO2, Kyoto, RGGI) • Capacity to drive short term and long term reductions in this context? • Given uncertainties, the key for understanding the market is the dynamics provided by the three flexibility mechanisms

Observed spot and DEC12 price over 2005-2012

5

I. Methodology

Counterfactual or baseline emissions

Reduction costs Initial distribution

of allowances

Behavior of participants with Flexibility mechanisms

Use of offsets

Banking Borrowing

Analysis of the past Prospective

exercise

Trading

6

Outline

Part I - Emissions trading: the key role of flexibility Part II - A prototype of ZEPHYR-Flex based on abatement and trading Part III - The use of carbon offsets: Good or Evil? Part IV - The calibration of ZEPHYR-Flex and the results on the first two trading periods Part V - Looking ahead: a multi-level regulation challenge

7

II. Baseline emissions

70

75

80

85

90

95

100

105

110

115

120

19

90

19

91

19

92

19

93

19

94

19

95

19

96

19

97

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

EU ETS industrial production (calculated from Eurostat, 2005=100)

CO2 Emissions EU27 (UNFCCC 1A1 and 1A2 sectors, 2005=100)

GDP EU27 (Eurostat 2005=100)

•Emissions as they would be without carbon policies • Scenarios based on the past relationship between GDP growth and emissions • Using production indexes as a sectoral breakdown of gross production rates • And an elasticity of emissions to production (0.6 in general; 1.2 in case of economic choc) Limit: Not differentiated by sector

Relationship between growth and emissions

60

70

80

90

100

110

120

2005 2006 2007 2008 2009 2010 2011 2012e

Electricity production

Rest of combustion

Refineries

Iron and steel

Cement

Rest of industry

EU ETS weighted index

Production growth by sector used for sectoral baselines

0

10

20

30

40

50

60

70

80

90

100

0% 10% 20% 30% 40%

Car

bo

n P

rice

(€

/tC

O2

)

Reductions below the baseline

EU ETS

?

8

II. Abatement

• Hypothesis on abatement costs of participating entities, by sector

• Cost can evolve over time

Relationship between price and abatement EU ETS level

Price

Quantities

Marginal abatement cost curve of a given installation

PEUA

EpEUA E0

Total cost of emissions reduction

Opportunity cost(the value of emission reductions at the market price is greater than the internal reduction costs)

E0-EpEUA: reduced emissions at price PEUA

• In the model, abatement is done instantly • Perfect recognition of opportunity cost • A share of reductions are removed from the emission baseline the following years

9

II. Results with trading only

0

5

10

15

20

25

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

€/t

CO

2

Carbon Price

Carbon price (Phase 1) Carbon price (Phase 2)Bluenext EUA 2005-07 Bluenext EUA 2008-12

1 500

1 600

1 700

1 800

1 900

2 000

2 100

2 200

2 300

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

Mill

ion

s

Total cap, baseline and verified emissions

Simulated emissions

Baseline Emissions

Cap (free+auction)

CITL Verified Emissions

Total cap (free+auction+offsets)

With trading only (no banking, no borrowing, no offsets): • Does not replicate observations • Despite the existence of a net surplus, trades are necessary for short installations to be compliant • Trading is very efficient to lower compliance costs price is zero as soon as baseline emissions are below the cap (assumption of zero transaction costs in the model)

10

Outline

Part I - Emissions trading: the key role of flexibility Part II - A prototype of ZEPHYR-Flex based on abatement and trading Part III - The use of carbon offsets: Good or Evil? Part IV - The calibration of ZEPHYR-Flex and the results on the first two trading periods Part V - Looking ahead: a multi-level regulation challenge

11

III. Offsets: Good or Evil?

• Major characteristic of the EU ETS: articulation with the Kyoto Protocol’s flexibility mechanisms, with quantitative and qualitative restrictions • The attitude towards offsets has changed: initially worshiped, Kyoto offsets are now perceived as aggravating “disequilibrium” on the EUA market • In the thesis:

• Observation of prices • Use of offsets at the installation level Effect on past EUA equilibrium price with ZEPHYR-Flex Scenario for the use of offsets up to 2020

12

III. Price evidence

• Existence of spot discount depends on whether the limit is binding or not • Different time discounts if offsets and allowances are not perceived as fully equivalent for future use

0

20

40

60

80

100

120

0 200 400 600 800 1000 1200 1400 1600 1800 2000

Pri

ce/M

argi

nal

Co

st

Emissions/Abatementy = -0,0395x

R² = 0,233y = -0,0365xR² = 0,6049

y = -0,0391xR² = 0,6161

y = -0,0383xR² = 0,5132

-25%

-20%

-15%

-10%

-5%

0%

5%

10%

15%

0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00

Ln(F

ut/

Spo

t)

(T-t) in years

DEC08 DEC09 DEC10 DEC11 DEC12 DEC13Linear (DEC08) Linear (DEC09) Linear (DEC10) Linear (DEC11) Linear (DEC12)

Announcement of

Phase 3 qualitative

restrictions

13

III. Observed and forecasted use

0

50

100

150

200

250

300

350

400

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

20

17

20

18

20

19

20

20

Mill

ion

to

nn

es

Offsets used in the EU ETS ZEPHYR-Flex scenario

2

0

0

8

2

0

0

9

2

0

1

0

2

0

1

1

2008 2009 2010 2011Average use of

annual limit

Constant users 83% 73% 92% 98% 86%

All except 2011 108% 118% 127% 118%

All except 2010 118% 96% 128% 114%

86% 70% 78%

All except 2009 91% 148% 147% 129%

83% 151% 117%

85% 182% 133%

Only in 2008 133% 133%

All except 2008 116% 115% 104% 112%

127% 71% 99%

95% 109% 102%

Only in 2009 135% 135%

171% 186% 179%

Only in 2010 281% 281%

Only in 2011 256% 256%

Absent users 0%

Average intensity of offsets use, by category of behavior etc…

+ Assumptions : •Phase 3 rules •Phase 3 offset supply •Evolution of behavior

Simulated use of offsets up to 2020

14

III. Lessons from Linking

• Decentralized rules can lead to more uncertainty on the cap, and thus on future price and emission trajectory • Legacy from the Kyoto framework: relying on outside standards can raise regulatory issues • Applying ex post restrictions: intervention on the rules/criteria can have an opposite effect in the short term

0

5

10

15

20

25

30

35

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

€/t

CO

2

Simulated Carbon price (Phase 1) Simulated Carbon price (Phase 2)

Bluenext spot price 2005-07 Bluenext Spot price 2008-12

• Clearly an arbitrage by participants on the market, leading to a relatively high use of offsets (550Mt in four years), and lower compliance costs • Nevertheless, economic optimum not fully established

• information problems • reputational concerns • transaction costs

15

Outline

Part I - Emissions trading: the key role of flexibility Part II - A prototype of ZEPHYR-Flex based on abatement and trading Part III - The use of carbon offsets: Good or Evil? Part IV - The calibration of ZEPHYR-Flex and the results on the first two trading periods Part V - Looking ahead: a multi-level regulation challenge

16

IV. Complete version of the model

We add the time flexibility into ZEPHYR-Flex: the banking and borrowing provisions Time flexibility mechanisms implies to represent the anticipations of participants which play an essential role in our calibration exercise. Three decision criteria (see behavior table in Annex):

1/ Present internal position (EUA Stock – Baseline Emissions)

Perfectly known

2/ Anticipated future internal position (Expected growth compared to free allocations) over an anticipation period

Imperfectly anticipated

3/ Market position through anticipation of bullish or bearish price Random (parameterized)

17

IV. Results after calibration

0

5

10

15

20

25

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

€/t

CO

2

Simulated Carbon price (Phase 1) Simulated Carbon price (Phase 2)

Bluenext spot price 2005-07 Bluenext spot price 2008-12

0

500

1 000

1 500

2 000

2 500

3 000

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

Mill

ion

to

nn

es

Verified EmissionsBaseline EmissionsCap (free+auction)CITL Verified EmissionsTotal cap (free+auction+offsets)

0

5

10

15

20

25

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

Mill

ion

to

nn

es

Borrowed Hedged EUAs Borrowed Compliance EUAs

0

500

1 000

1 500

2 000

2 500

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

Mill

ion

to

nn

es

Banked Hedged EUAs Banked Compliance EUAs

18

• The EU ETS induced a cumulated amount of emission reduction of about 1,540 Mt over the first two Phases. 60% have been obtained outside the scheme’s perimeter through the use of carbon offsets (900 Mt) • The total compliance costs are estimated at 30 bn€ (reductions, offsets and auctions); exchange of allowances between participants represents a value of 18 bn€ over the first two phases • Given the level of the allowance cap, the cumulated net banking is close to 2,000 Mt at the end of 2012 Anticipations are preventing the price from being zero

Dynamic of behavior in the future ? (shift to auctions, change of behavior following an intervention from public authority…)

IV. Results after calibration

19

Outline

Part I - Emissions trading: the key role of flexibility Part II - A prototype of ZEPHYR-Flex based on abatement and trading Part III - The use of carbon offsets: Good or Evil? Part IV - The calibration of ZEPHYR-Flex and the results on the first two trading periods Part V - Looking ahead: a multi-level regulation challenge

20

V. Baseline with policy interactions

Renewable Energy and Energy Efficiency policies can affect EU ETS perimeter’s emissions independently from the EUA price: Integration in the model as a progressively lowered elasticity between production growth and baseline emissions growth (average of 60 Mt/yr in Phase 2 to 100 Mt/yr in 2020)

EMISSIONS CAP 30% BELOW BAU

BAU EMISSIONS

Reductions from: energy efficiency policestechnology policiesprice response in trading scheme

10 %

15 %

SUPPLEMENTARY POLICIES UNDERACHIEVE

(a)

5 %

SUPPLEMENTARY POLICIES OVERACHIEVE

(b)

Source: Baron (2012)

21

V. Simulation of interventions

Three intervention scenarios are tested with the model:

• The introduction of a reserve price of 20€/t at auctions • A back-loading (or set-aside) of allowances as proposed by the Commission on July, 25th 2012 • A reevaluation of the allowance cap to 2020 and 2030 compatible with the Roadmap 2050

0

5

10

15

20

25

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

20

17

20

18

20

19

20

20

€/t

CO

2

0

5

10

15

20

25

30

35

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

20

17

20

18

20

19

20

20

€/t

CO

2

Set aside 400Mt Set aside 900MtSet aside 1200Mt

0

5

10

15

20

25

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

20

13

20

14

20

15

20

16

20

17

20

18

20

19

20

20

€/t

CO

2

Roadmap (no anticipation change)

Roadmap (change of anticipation)

22

V. Lessons

• Some intervention measures can induce more uncertainty and instability, as they would not allow firms to set up “correct” anticipations by themselves • It is very difficult to send the “right” incentives to market players in the absence of explicit long term targets that are connected with the current and medium term cap (US SO2 trading program included a 30 years cap) • The unforeseen evolution of the macro-economic context, the state of international carbon markets and the link with an offsetting mechanism are factors which can be extraordinarily stimulating but also undermining if not dealt with “appropriately”

23

Limits, and how to extend this work

Abatement • The abatement costs curves by sectors, and their potential shift over time • The effect of abatement on the baseline, and on cost curves: short term versus long term reductions, innovation effect…

Baseline emissions

• Sector elasticities • Interactions with other policies • Explicit link with relevant drivers (energy prices, country growth etc.)

Compliance and hedging behavior

• Transaction data (banking, borrowing, maybe more) • Identification to a decision model ?

Modeling and Linking other markets • Aviation, Australia, California ?

Thank you for your attention

October 17th, 2012 – Paris Dauphine University (D520)

Climate Economics Chair - raphael.trotignon@chaireeconomieduclimat.org

mailto:raphael.trotignon@chaireeconomieduclimat.org

25

Working papers and peer reviewed articles

• Trotignon, R. and Delbosc, A. (2008), Allowance Trading Patterns During the EU ETS Trial Period: What does the CITL reveal?, Climate Report n°13, CDC Climat Research, June 2008 • Ellerman, A. D. and Trotignon, R. (2009), Cross Border Trading and Borrowing in the EU ETS, The Energy Journal, Volume 30 (Special Issue 2), Climate Change Policies After 2012, 2009 • Trotignon, R. and Leguet, B. (2009), How Many CERs by 2013?, Mission Climat of Caisse des depots, Working Paper n°2009-5, August 2009 • Trotignon, R. and Solier, B. (2011), The European Market on the Road to Phase 3, in De Perthuis, C. and Jouvet, P-A., eds. (2011), Climate Economics in Progress 2011, Chapter 2, Economica, October 2011 • Trotignon, R (2012), Combining cap-and-trade with offsets: lessons from the EU-ETS, Climate Policy, Vol. 12, Iss. 3, 2012 • De Perthuis, C. and Trotignon, R. (2012), The European CO2 allowances market: issues in the transition to Phase III, Climate Economics Chair, Information and Debates n°14, March 2012

26

Complete version of ZEPHYR-Flex

Total Allowances for Year Y (annual cap)

MACCs

Market exchanges: Supply/Demand equilibirum for Year YThe price starts from zero and goes up as long as ∑Buy +/- ∑Hedging-Spec > ∑Sales + ∑Auctions + ∑Offsets

Verified EmissionsYear Y

Free Allocation Auctions

Installation 1 Installation 2 Installation 10,000Installation 3 …

EUA StockBaseline

Emissions

AnticipatedFuture price

Buying Selling Banking Borrowing Hedging/Speculation

Position for Year Y

Anticipated position over the period

Year N+1

Use of Kyoto offset for Year Y (exogenous)

Emission ReductionsYear Y

EUA equilibrium price Year Y

EUA Price

Bought/SoldYear Y

Banking/BorrowingYear Y