Connect with Vanguard > vanguard.com

Executive summary. The global commercial real estate market has been estimated to be as large as $23 trillion.1 Historically, commercial real estate has provided competitive real returns and diversification opportunities for traditional portfolios. Yet an important question remains: Can an investment in commercial real estate actually deliver the characteristics and benefits of the broad real estate market? Indeed, investment vehicles such as real estate investment trusts (REITs), partnerships, or private investment pools can look quite different from the broad real estate market. The complexity of this question is a possible reason why institutional investors on average allocate only 6% of their portfolios to commercial equity real estate (Cerulli Associates, 2011). In the United States, as of December 2010, private commercial real estate holdings were estimated at $247 billion, according to the National Council of Real Estate Investment Fiduciaries (NCREIF), and public U.S. equity REITs at $358 billion.2 This analysis evaluates the commercial real estate market and offers perspective regarding the various investment options.

Vanguard research March 2011

1 Estimated market size based on data provided by Prudential Real Estate Investors.2 For this paper, we use real estate investment trust or REIT to refer to any form of equitized real estate investment.

REITs: Effective exposure to commercial real estate?

Authors

Christopher B. Philips, CFA

David J. Walker, CFA

Yan Zilbering

2

Notes on risk: All investments are subject to risk. Funds that concentrate on a relatively narrow market sector face the risk of higher share-price volatility. Foreign investing involves additional risks, including currency fluctuations and political uncertainty. Past performance is not a guarantee of future results. The performance of an index is not an exact representation of any particular investment, as you cannot invest directly in an index. Diversification does not ensure a profit or protect against a loss in a declining market.

We contend that:

• Commercialrealestateisauniqueandsignificantassetclass.

• Anequitizedrealestateindexservesasalong-termproxyforthecommercialreal estate market.

• BecauseREITsrepresentexposuretothecommercialrealestateassetclass,theymaybeevaluated for strategic inclusion in a portfolio using the same metrics employed for other asset classes.

• However,becauseREITsareequitiesandarethereforepartofbroad-basedglobalequitymarkets, investors must factor in the exposure already contained within the active and indexed portions of their portfolios.

Commercial equity real estate basics

Commercial equity real estate is a unique and significant asset class. It provides investors with the opportunity to own property, with the primary objective of leasing a structure to various tenants in return for income and the potential for capital gain upon the sale of the property.

Commercial real estate may be segmented by property type, geographic location, and development stage. Core property types commonly include multifamily residential (apartment buildings), retail, office, and industrial. Other, less common property types include health care facilities, public storage buildings, golf courses, and hotels. Development stagesincludecore(mature,income-producingproperties),value-added(income-producingproperties with appreciation potential), and opportunistic or distressed (development properties with a focus on capital appreciation).

Long-termreturnsforrealestatearedrivenprimarilyby a property’s net operating income (NOI) and to a lesser extent by property value appreciation. Historically, commercial real estate prices and income (assuming properties are maintained appropriately) have grown on par with to slightly above inflation (Figure 1,page4).Income-orientedproperties may increase rents to compensate for inflation’s erosion of value, and thus obtain some hedge against inflation. However, for a property to maintainitsinflation-adjustedvalue,itmustbewell-located, have rents that can be adjusted periodically, and not be subject to sudden, sharp increases in operating costs. As an extreme example, if an apartmenthadrent-controlprovisions,theproprietorwould not be able to increase rents, and the property therefore would not provide an inflation hedge.

3

A real estate investment trust is an operating company that offers ownership shares to the public and, as a result, trades on a stock exchange. Equity REITs own property directly and derive most of their earnings from property income, with a smaller portion coming from appreciation when an underlying holding is sold. As a consequence of various tax and income requirements,3 many equity REITs choose to investprimarilyinmature,income-generatingproperties. Mortgage REITs make loans to property developers, and hybrid REITs combine the two strategies. In this report we use equity REITs as a lens for the market, because the othertypesdonotexplicitlyfocusonincome-producing properties. We thus employ the FTSE NAREIT Equity REIT Index as a proxy for U.S. REIT investments.4

Private real estate investment can take the form ofdirectinvestments,open-orclosed-endedcommingled funds, or private partnerships. In a direct investment, an investor purchases and managesapropertydirectly.Closed-ended

commingled funds represent large pools of capital, with the underlying properties managed byaspecialist.Open-endedcommingledfundsoffer investors more freedom to add or withdraw capital—given proper notification and the approval ofthemanager.Bytheirverynature,privatepartnershipsarehighlycustomizable.Withouttaxlaws to govern their structure, private partnerships can focus on any combination of property types and development stages.

The most popular benchmark used to measure private investment holdings is the NCREIF PropertyIndex.Becauseofwell-documentedconcernsabouttheappraisal-basedvaluationmethodology, which introduces serial correlation and return (volatility) smoothing, we elected to useatransactions-basedindexthatwasdeveloped by the MIT Center for Real Estate (Fisher, Geltner, and Pollakowski, 2006) and endorsedbyNCREIF.Thetransactions-basedindex corrects for smoothing, leading to standard deviations significantly greater than reported in the standard index.

Public and private real estate investments

3 Tax laws mandate that a REIT must pay out 90% of its taxable profit in the form of dividends. In addition, a REIT must continually satisfy certain tests with respect to the sources of its income, the nature and diversification of its assets, the amount of its distributions to stockholders, and the ownership of its stock. Among other things, these restrictions may limit the company’s ability to acquire certain types of assets, limit its ability to dispose of assets that it has held for less than four years if the disposition would result in gains exceeding specified amounts, limit the ability to engage in hedging transactions, and require the company to make distributions to its stockholders at times when the company might deem it more advantageous to use those funds for other corporate purposes or when the company might not have funds readily available for distribution.

4 For analysis using the FTSE NAREIT Equity REIT Index, many choose to focus on the period starting in the early 1990s, because of the significant tax and structural changes that followed the 1986 Tax Reform Act. The act permitted REITs to operate and manage most types of income-producing commercial properties by providing “customary” services associated with real estate ownership. Between 1986 and 1992, real estate suffered a severe recession, and by 1992 many private real estate companies realized that the most efficient way to access capital was from the public marketplace through REITs (source: www.reit.com).

Sustainable NOI is used to calculate the current and future value of a property. Expectations of future NOI (and hence property values) change with economic cycles, the supply and demand of properties and tenants, and the performance of the financial markets (stocks and bonds compete with real estate for investor capital).

Investing in commercial real estate

Traditionally, there are three primary means of investingincommercialrealestate,assummarizedin Figure 2, page 4: direct investment and management, participation in a private investment pool, and purchases of REIT shares.

4

5 Although illiquidity potentially leads to a return premium in most assets, whether or not such a premium exists for an investment pool is the subject of an ongoing debate in the investment industry.

Direct investment involves purchasing a property outright and assuming operating control over rent policy and collection, maintenance, and growth. Although direct ownership allows the investor to collect rents and the proceeds of any property sales without paying a management fee, it requires expertise in property management, or a willingness tohireathird-partymanager.Directownershipalsolimitstheabilitytocreateamulti-propertyportfoliobecauseofthesizeoftherequiredinvestments.Becauseofthesechallenges,fewdiversifiedrealestate investors choose to own properties directly.

Private real estate vehicles provide direct access to commercial real estate properties. However, for many investors, private real estate can be difficult to include in a diversified portfolio because of high costs, illiquidity,5 limited transparency, and large requiredminimuminvestments.Andevenanopen-

Figure 1. Title of chart goes here

Figure 1.Title of chart goes herethree lines chart titlethree line chart title

Figure 1. Title of chart goes heretwo lines of chart title

Figure 1. Title of chart goes here

Figure 1.Title of chart goes herethree lines chart titlethree line chart title

Figure 1. Title of chart goes heretwo lines of chart title

45%35%15%5%

Slice 1Slice 2

Real estate prices Rental incomes

Slice 3Slice 4

Line 1Line 2Line 3Line 4

Source: Author’s calculations using data from the Bureau of Economic Analysis and Moody’s Analytics. The BEA updates data in August of each year.

0.00

0.25

0.50

0.75

1.00

1.25

1.50Annualized Real Growth in Real Estate Rental Incomes

Annualized Real Growth in Real Estate Prices

0.0%

0.5%

1.0%

1.5%

Figure 1. Inflation-adjusted growth in national nonresidential real estate prices and rental income, 1947–2009

Ann

ualiz

ed r

eal g

row

th

Direct investment Private investment pool REIT shares

Form of investment Purchase investment property directly

Invest in commingled trust or private partnership

Purchase shares on stock exchange

Management Direct management of property Investment manager allocates funds and manages property portfolio.

Investment management company (REIT) manages property portfolio.

Investor time and resource requirements

High Variable Low

Source of returns Rents received; property value appreciation

Pooled rents; partnership share appreciation

Dividends paid; share price appreciation

Valuation Annual appraisal; income capitalizationrate

Pooled appraisals; pooled capitalizationrates

Funds from operations (FFO) multiples; NAV/market cap

Return characteristics Highly idiosyncratic Idiosyncratic—highly subject to manager skill

Can range from idiosyncratic (single REIT) to systematic (REIT index)

Liquidity Highly illiquid Illiquid—subject to purchase and redemption restrictions and transaction minimums

Highly liquid

Management fees None Higher Lower

Diversification? No Limited Yes

Transparency Transparent (property owner) Little portfolio transparency Transparency consistent with requirements for publicly traded companies

Characteristics of primary real estate vehiclesFigure 2.

Note: There may be other material differences between products that must be considered prior to investing.Source: Vanguard.

5

ended commingled fund often requires early notification and substantial waiting periods for account transactions. Finally, to attain diversification across property types and regions, the investor needs to either buy a share in a vast portfolio or acquire shares in multiple portfolios.

In terms of the underlying real estate exposure, there is no substantive difference between holding an interest in a private real estate partnership and holding a REIT. However, ownership of REIT shares addresses many of the investment concerns associated with private real estate partnerships or pools. REITs offer transparency and liquidity far exceeding that of most private real estate investments.Andregionalandproperty-typediversification is achieved more easily in public real estate than in private real estate. For example, a REIT index fund may hold more than 100 REITs, each representing many underlying properties across geographic regions.

Are REITs an effective proxy for the real estate market?

Public real estate options come with their own unique concerns. Given the desirable investment qualities of REITs, for many investors the primary question is whether REITs can provide effective exposure to real estate. Although REITs constitute a small portion of the real estate market and have performed quite differently from other real estate vehicles in the short term, we believe there are at least two fundamental similarities leading to the conclusion that REITs are representative of the commercial real estate market. These are geographic representationandlong-termperformance.

Most important, REITs and private investments— whether held directly or through a private investment pool—are invested similarly. Indeed, both private and public investment pools derive returns from holding portfolios of commercial real estate. As Figure 3 suggests,theREITmarketiswell-diversifiedacrossproperty types, including a significant portion of the property types that are likely to be included in a private investment pool. In addition, a REIT index is geographically diversified, representing holdings from all areas of the country. The range of property types

and geographic regions suggests that a broad REIT index is more representative of the aggregate real estate market than any single REIT or private investment pool.

It is also instructive to examine the performance of REITs relative to that of private holdings. Given that investment properties represent the underlying holdings of both public and private investment vehicles, returns should be more similar than different, particularly over the long term. However, because the primary public and private benchmarks have modestly different constituents (the NCREIF Property Index includes office, retail, industrial, and residential properties, along with a very small percentage of hotels, and includes properties across the range of development stages), we would expect some differences in returns, particularly in the short term. To help illustrate this concept, Figure 4, page 6, shows that since 1984 (the inception of the NCREIF transaction index), the cumulative returns of public and private real estate have not been meaningfully different over longer periods when private real estate is adjusted for differences in how benchmark returns are reported; see the note below Figure 4 for details. From this

Sector composition of FTSE NAREIT Equity REIT Index

Figure 3.

As of December 31, 2010

Note: Percentages may not sum to 100% because of rounding.

Source: FTSE/NAREIT.

Apartments 16%

Regional malls 15%

Offices 14%

Health care 12%

Shopping centers 9%

Lodging/resorts 6%

Diversified 6%

Specialty 6%

Self-storage 6%

Industrial 5%

Office/industrial mix 3%

Free-standingretail 2%

Manufactured homes 1%

6

perspective, investors in public or broad private real estate indexes would have ended up in essentially the same place over longer periods of time, albeit by marginally different routes.

Although the use of publicly held real estate vehicles as a proxy for the commercial real estate market seems reasonable, it is clear that returns for public REITs and private pools may differ significantly in the short term, as seen in Figure 5. Here we show that overthree-yearperiods,theFTSENAREITandNCREIF transaction indexes have performed quite differently. This could be due to differences in holdings or, more important, to a potentially significantshort-termrelationshipbetweenREITsand public equity markets.

Arguments against using REITs

Asaresultofthepotentialforshort-termperformance disparity, the primary argument against using REITs as a proxy for the commercial real estate market is the correlation of REIT performance with the broad equity market—particularly in the small-capvaluesector.HigherhistoricalcorrelationsbetweenREITsandsmall-capvaluestocksdosuggest that some return variation is related to the movementsofthesmall-valuesector.However,asignificant portion of REITs’ return variation (approximately 50%) is uncorrelated, indicating substantial independence.

The primary differences between REITs and traditional equities are the core businesses and the primary drivers of earnings growth. REIT earnings are driven by the net operating income of the property holdings and, to a lesser degree, by appreciation in the value of properties. In contrast, traditional corporate earnings are driven by growth in sales revenue for products and services. As a result of this difference in business

Figure 4. Long-term returns are similar across real estate investment options

Cumulative returns for December 1983–September 2010

NCREIF MIT transactions-based index with 50% adjustmentEquity REITs

Note: There are two adjustments to the traditional NCREIF Property Index. First, we use the transactions-based index developed by the MIT Center for Real Estate and provided by NCREIF. This eliminates the smoothing resulting from appraisal pricing. Second, we adjust the NCREIF returns to account for the discrepancy with which returns are reported. Because REIT returns are a function of capital structure, the return series accounts for any debt/equity ratio the firms employ. In contrast, private real estate returns assume that property investment is made without the use of debt financing. In fact, returns represent changes in property value only, not the returns realized by investors who may be partially financed with debt. Academic and industry analyses typically include adjustments to private real estate returns of anywhere from 30% to 70%. Our adjustment is an attempt to make the return series more comparable, with the acknowledgment that there are perhaps more complex and more accurate approaches.

Source: Author’s calculations using data from FTSE/NAREIT and from NCREIF with MIT. Analysis is derived from the methodology employed by researchers such as Joseph Pagliari Jr. (2003). The transactions index data is released on a quarterly lag.

0

500

1000

1500

2000

2500 NCREIF MIT Transaction Index: 50% Adjustment

Equity REITs

Sept. 2010Dec. 2007Dec. 2003Dec. 1999Dec. 1995Dec. 1991Dec. 1987Dec. 1983

0

500

1,000

1,500

2,000

2,500

Sept. 2010Dec. 2007Dec. 2003Dec. 1999Dec. 1995Dec. 1991Dec. 1987Dec. 1983

Cum

ulat

ive

perf

orm

ance

in

dexe

d to

100

in D

ecem

ber

1983

NAREIT Index

NCREIF MIT Index

NCREIF MIT Index with 50% adjustment

Mean quarterly return 3.00% 1.93% 2.90%Statistically different from REITs?

No: t-stat=–1.045

No: t-stat=–0.094

7

Figure 5. Shorter periods can result in significant performance differences

Figure 1. Title of chart goes here

Figure 1.Title of chart goes herethree lines chart titlethree line chart title

Figure 1. Title of chart goes heretwo lines of chart title

Figure 1. Title of chart goes here

Figure 1.Title of chart goes herethree lines chart titlethree line chart title

Figure 1. Title of chart goes heretwo lines of chart title

Total cumulative returns: Three-year nonoverlapping periods

45%35%15%5%

Slice 1Slice 2

FTSE NAREIT Equity REIT IndexNCREIF MIT transactions-based index with 50% adjustment

Slice 3Slice 4

Line 1Line 2Line 3Line 4

Source: Author’s calculations using data from FTSE/NAREIT and from NCREIF with MIT. Transactions-based index data are released on a quarterly lag. The final bars reflect data through September 30, 2010. See the note to Figure 4 for an explanation of the 50% adjustment.

–40%

–20%

0%

20%

40%

60%

80%

100%

120%

2008–9/20102005–20072002–20041999–20011996–19981993–19951990–19921987–19891984–1986

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2NCREIF MIT Transactions Index with 50% adjustment

FTSE NAREIT U.S. Equity REITs

2008-9/20102005-20072002-20041999-20011996-19981993-19951990-19921987-19891984-1986

fundamentals, there can be other significant contrasts betweentraditionalsmall-capvaluestocksandREITs,as illustrated in Figure 6.

A second argument commonly used against REITs is thatequitizationofrealestateexposuremaynotwork because it hasn’t worked with other asset classes—mainly commodities. The impact of equitizationisoftenevaluatedinthecommoditymarkets, where equity sectors such as those

focused on energy or mining firms can be used as proxies for a commodity index itself. However, investing in real estate typically centers on a decision about whether to obtain exposure through a publicly listed REIT or a privately held partnership—that is, a choice between public and private vehicles, not a decision about whether to hold property outright or as a share in a managed portfolio of properties. The equitizationofcommoditiesinvolvesdecidingwhether to invest in the commodity itself (say, a

REIT fundamentals differ from those for small-cap value stocksFigure 6.

Small-cap value REITs

Core business/earnings Traditional revenue sources Real estate management

Balance-sheetassets Depreciating Appreciating

Source of returns: Price 8.53% 3.41%

Source of returns: Income 2.29% 6.91%

Annualizedvolatility 17.16% 19.14%

Average dividend yield 2.17% 6.54%

Sources: Russell and FTSE/NAREIT. Data reflect the period from 1989 through 2010; 1989 represents the start of the Russell 2000 Value Index dividend yield data.

8

gold bar), in commodity futures (the gold component of a commodity index), or in shares of companies whose primary business is the production of that commodity(gold-miningcompanies).

Commercial real estate market exposure

For investors in any asset class to obtain the risk and return characteristics of the market, an investable market index is required. However, there are two primary concerns with commercial real estate indexes. First, existing benchmarks (such as the FTSE NAREIT and NCREIF indexes) represent only a small portion of the total U.S. commercial real estate market. For example, the total value of U.S. commercial real estate is likely measured in trillions; since investment holdings account for approximately $600 billion, it appears that the vast majority of the commercial real estate market is closely held by real estate developers, brokers,andlong-terminvestorsandisthereforeunavailable for transaction. This is somewhat different from the equity and debt markets, where public holdings make up a much larger portion of the total marketcapitalizationofeachassetclass.

Second, only REIT indexes have historically been investable, as the NCREIF Property Index represents an aggregation of private investor holdings. In fact, outside of a REIT index mutual fund or ETF, any investment in commercial real estate is highly concentrated and consequently a bet on manager skill. Investment performance therefore entirely depends on the selection of a manager and that manager’s skill.

Thedifferencebetweenspecializedexposureandmarket exposure is important, because investors typically model asset allocation and expected portfolio risks and returns on the basis of broad market exposures. In other words, investors put the characteristics of a particular asset class into a model to determine an appropriate allocation. Many asset classes, such as domestic and international stocks and bonds, can be replicated to capture the market exposure,butspecializedinvestmentssuchasprivately held commercial real estate do not yet offer

this option to the same degree. The derivatives markets may eventually help investors to track private real estate indexes, but for now such opportunities are limited. The importance of manager selection therefore results in a potentially wide distribution of return possibilities. In fact, if investors in private vehicles receive the return for the overall real estate market, it is more likely to be because of luck than because they found an investment pool that is diversified enough to replicate the true characteristics of the aggregate real estate market.

Figure 7 demonstrates the relationship between idiosyncratic risk and systematic risk. Here we plot five-yearannualizedreturnsofindividualREITsovertime, with each dash representing the return for a specific REIT.6 For example, in 1993, 46 REITs had five-yearreturns,whilein2009,141hadfive-yearreturns. The dispersion of the returns simulates the risk inherent to selecting a concentrated real estate portfolio. We also plot the index return as well as the 25th,50th,and75thpercentilereturns.Bydefinition,in any given period, 50% of firms will be above the median and 50% will be below the median, but it is instructive to look at the range of returns as well as where the index falls relative to the distributions. In everyfive-yearperiod,apercentageoffirmspostedreturns well below “average.”

It is also instructive to evaluate the spread between the 25th and 75th percentiles. In most years this spread was between 15 and 20 percentage points. Such a wide spread sheds light on the challenges associated with using undiversified investment vehicles to implement an asset allocation strategy. In this example, investors can own the REIT market, which means that the decision to invest with a manager or through an index can be a part of the investment process. However, no such investable index exists for the remainder of the commercial real estate market. As a result, investors in private partnerships must be able to believe that their managers can consistently deliver returns that are not only above the median return for all private partnerships but also above the index return for public REITs.

6 We use REIT returns here because, unlike return data for private partnerships, REIT data are accurate and free of biases. Further, a REIT can be thought of as a publicly traded version of a private partnership or private investment pool. The investment managers invest in similar properties, with similar goals, objectives, and risks.

9

Figure 7. The dangers of concentrated investment in commercial real estate

Figure 1. Title of chart goes here

Figure 1.Title of chart goes herethree lines chart titlethree line chart title

Figure 1. Title of chart goes heretwo lines of chart title

Figure 1. Title of chart goes here

Figure 1.Title of chart goes herethree lines chart titlethree line chart title

Figure 1. Title of chart goes heretwo lines of chart title

The distribution of five-year annualized returns for individual REITs, 1988–2010

45%35%15%5%

Slice 1Slice 2Slice 3Slice 4

Line 1Line 2Line 3Line 4

25th percentile50th percentile75th percentileFTSE NAREIT All REITs Index

Source: Author’s calculations using data from FTSE/NAREIT and Morningstar.

-60

-40

-20

0

20

40

60

-60

-40

-20

0

20

40

60

–60%

–40%

–20%

0%

20%

40%

60%

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

Portfolio impact of equitized real estate

To evaluate the potential impact of real estate exposure on a portfolio, we used the FTSE NAREIT Equity REIT Index to approximate the real estate market. This presents an obvious challenge: We have provided evidence to support the idea that commercial real estate is a separate asset class, naturallysubjecttoasset-allocationdecisionssimilarto those made for stocks and bonds. Further, we believe that REITs represent the best opportunity for many investors to gain access to this asset class. The waters are muddied, however, because while REITs represent exposure to a unique asset class, they are in fact equities.

BecauseREITsareequities,tojustifytheirinclusionina financial model as a unique asset class, investors must believe that REITs offer an exposure fundamentally different from that provided by other

equities—a case that is reasonably strong. For example, the returns of most equities may be almost entirely explained by three key variables—the returns ofthebroadequitymarket,asizefactor(largeorsmall), and a style factor (growth or value). These are commonlyknownastheFama-Frenchriskfactors.However, returns for REITs are only partially explained by these fundamental factors. In fact, the underlying real estate holdings constitute the largest single influence on REIT returns.7 It can be contended, therefore, that investors gain access to the previously unavailable real estate market through REITs.

Of course, while arguments can be made regarding the utility of adding a unique REIT allocation to a portfolio, investors must account for the fact that REITs are equities and as such are often contained in their broad equity allocations.8 In fact, for many investors, this may represent the only commercial real estate exposure required.

7 Separately we find that the Fama-French factors explain only 50% of REIT returns, leaving a large error term, likely representative of the real estate asset.8 Investors with heavy allocations to active equity managers will be less able to control the total allocation to REITs, as the underlying managers themselves

may be overweighting or underweighting REITs within their specific portfolios. In fact, a given strategic allocation is known for certain only if the entire portfolio is indexed.

10

Overall, to advocate a specific allocation to REITs, an investor must first believe that a broad REIT index fund or ETF accurately represents systematic exposure to commercial real estate. If the investor does not believe that but still desires exposure to commercial real estate, then an allocation to private real estate requires confidence that the private manager will consistently produce returns above those achieved by the median private manager as well as by the REIT index.

If an investor does accept that an index vehicle provides systematic exposure to the commercial real estate market, then the investor must also be comfortable with potentially significant deviations from other real estate indexes in the short term (Figure 5). And the investor must be confident that the portfolio will benefit from exposure above that provided by the inclusion of REITs in the public equity market. Assuming that these conditions hold for certain investors, then an independent strategic allocation to real estate may be warranted. Such strategic allocations to REITs should be based on specific risk and return objectives and constraints, and portfolios should be periodically rebalanced to those allocations.

Looking beyond the United States

So far in this analysis, we have made the case for commercial real estate using benchmarks focused on the U.S. market. This is primarily due to the abundance of data in the United States and the relative lack of data globally. That said, a case can be made that, as with equities and fixed income, investors thinking about including a real estate proxy in a broadly diversified portfolio should not limit themselves to a focus on one country or currency. Indeed, to the extent that global investment options are available, a portfolio’s diversification may be enhanced by including exposure to both U.S. and non-U.S.components.9

Atahighlevel,therationaleforincludingnon-U.S.real estate in a portfolio that already contains a U.S. allocation to this asset class is similar to the argumentforcombiningU.S.andnon-U.S.equities—namely, increased diversification and the potential for lower volatility. For example, global diversification permits exposure to multiple property markets, business cycles, and interest rate regimes. Imperfect correlations across these markets would theoretically lead to lower overall portfolio volatility.

Figure 8 reveals that as of December 2010, U.S. REITs accounted for only one third of the global equitizedrealestatemarket.ByfocusingsolelyonU.S. REITs, an investor is actively excluding two thirds of the global market from the real estate portfolio. It would therefore make intuitive sense that a U.S. investor could obtain a diversification benefitfromexpandingtheportfoliotoincludenon-U.S. REITs.

The left panel of Figure 9, page 12, shows that historically the U.S. REIT market has been imperfectlycorrelatedtothenon-U.S.realestatemarket. While correlations have been trending up since the early 2000s, the potential for diversification remains as long as correlations are less than 1.10

The right panel of Figure 9 reveals that, while the impact to portfolio volatility has cycled between higher and lower volatility, portfolios including both U.S.andnon-U.S.realestateexposurewouldhaveexperienced lower average volatility than a portfolio solely invested in U.S. real estate. This can be seen whereverthelineinthegraphdipsbelowthex-axis.It is notable that, despite the increasing correlations, for much of the 2000s a portfolio that incorporated global real estate exposure would have experienced lower average volatility, even during the financial crisis of 2008.

9 Outside the United States, the most prevalent structure for equitized real estate companies is a real estate operating company or REOC. A REOC is similar to a REIT, except that a REOC reinvests earnings into the business, while a REIT distributes earnings to the shareholders. In addition, REOCs may be able to make more types of investments than REITs can.

10 Interestingly, this is the same trend we have witnessed between U.S. and non-U.S. equities over the past decade (see, for example, Philips, 2011). However, the trend is stronger in the broad equity market than in the real estate market, suggesting a potentially greater diversification benefit from combining U.S. and non-U.S. real estate.

11

Conclusion

Commercial real estate is a unique and significant assetclass.Becauseofthis,anargumentcanbemade for its inclusion in a diversified portfolio. The challenge investors face, however, is that unlike equities, fixed income, or commodities, the available vehicles do not offer pure exposure to the asset class.WhetherutilizingREITs,acollectivetrust,aseparate account, or direct property ownership, investors are only exposed to a small slice of the broad commercial real estate market. As a result, real estate investors must be comfortable with the potential for their investment to deviate significantly from the performance of that broad market.

For those investors who desire exposure to commercial real estate and are indeed comfortable withthepotentialshort-termrisks,wehaveshownthat a broad REIT index can serve as an effective proxy for the real estate market. In other words, we believe it is unnecessary for investors to incur the illiquidity, high costs, and manager risk of a relatively concentrated privately managed portfolio of commercial properties.

Finally, while we have shown that REITs offer liquid, diversified,transparent,andlow-costexposuretocommercial real estate, investors must be comfortable with one last risk—the risk of a sector overweight. At the end of the day, REITs are equities, and as such are represented in most broadly diversified equity funds. For example, as of December 2010, REITs accounted for just over 1% of the broad U.S. stock market. The reality is that any additional allocation to REITs can represent a significant overweighting of a potentially volatile and concentrated sector.

Figure 1. Title of chart goes here

Figure 1.Title of chart goes herethree lines chart titlethree line chart title

Figure 1. Title of chart goes heretwo lines of chart title

Figure 1. Title of chart goes here

Figure 8. Makeup of the global real estate market

Figure 1.Title of chart goes herethree lines chart titlethree line chart title

Figure 1. Title of chart goes heretwo lines of chart title

S&P Global Property Index country weightings

45%35%15%5%

Slice 1Slice 2Slice 3Slice 4

Line 1Line 2Line 3Line 4

Source: Standard & Poor’s. Data are as of September 30, 2010.

0 5 10 15 20 25 30 35

South Korea

Russia

Portugal

Peru

Morocco

Mexico

Ireland

Hungary

Czech Republic

Turkey

Spain

Norway

Luxemborg

Italy

Greece

Denmark

Chile

Poland

New Zealand

Finland

Thailand

Egypt

Maylasia

Isreal

Indonisia

Belgium

Germany

Philippeans

Taiwan

Austria

Switzerland

Brazil

Sweden

India

Netherlands

South Africa

Canada

France

United Kingdom

China

Singapore

Australia

Japan

Hong Kong

United States

0 5 10 15 20 25 30 35%

Turkey

Spain

Norway

Luxembourg

Italy 0.10%

Greece

Denmark

Chile

Poland

New Zealand

Finland

Thailand

Egypt

Malaysia

Israel

Indonesia

Belgium

Germany

Philippines

Taiwan

Austria

Switzerland

Brazil

Sweden

India

Netherlands

South Africa

Canada

France

United Kingdom

China

Singapore

Australia

Japan

Hong Kong

United States

0.10%

0.10%

0.10%

0.10%

0.10%

0.10%

0.10%

0.20%

0.20%

0.20%

0.30%

0.30%

0.40%

0.40%

0.40%

0.40%

0.50%

0.60%

0.80%

0.90%

1.00%

1.00%

1.10%

1.10%

1.50%

1.60%

2.10%

4.20%

4.30%

4.80%

5.00%

8.30%

11.00%

12.80%

33.8%

12

Figure 9. Historically U.S. REITs and non-U.S. real estate have been imperfectly correlated

Figure 1. Title of chart goes here

Figure 1.Title of chart goes herethree lines chart titlethree line chart title

Figure 1. Title of chart goes heretwo lines of chart title

Figure 1. Title of chart goes here

Figure 1.Title of chart goes herethree lines chart titlethree line chart title

Figure 1. Title of chart goes heretwo lines of chart title

12-month rolling correlations between U.S. REITsand non-U.S. real estate

12-month rolling impact on portfolio volatility

July 1990–December 2010 July 1990–December 2010

45%35%15%5%

Slice 1Slice 2Slice 3Slice 4

Line 1Line 2Line 3Line 4

20% 30% 50%

Sources: Morningstar and Standard & Poor’s. U.S. REITs are represented by the S&P U.S. REIT Index. Non-U.S. real estate is represented by the S&P Global Ex U.S. Property Index.

2009200720052003200119991997199519931991

–0.6

–0.4

–0.2

0.0

0.2

0.4

0.6

0.8

1.0

2009200720052003200119991997199519931991

–15%

–10%

–5%

0%

5%

10%

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1/1/101/1/091/1/081/1/071/1/061/1/051/1/041/1/031/1/021/1/011/1/001/1/991/1/981/1/971/1/961/1/951/1/941/1/931/1/921/1/91-0.15

-0.10

-0.05

-0.00

0.05

0.10

1/1/101/1/091/1/081/1/071/1/061/1/051/1/041/1/031/1/021/1/011/1/001/1/991/1/981/1/971/1/961/1/951/1/941/1/931/1/921/1/91

Allocations to non-U.S. real estate

13

References

CBRichardEllis,2004.Property Investment Funds: On the REIT Road?

Center for International Securities and Derivatives Markets, 2005. The Benefits of Real Estate Investment: 2005 Update. CISDM Research Department.

Cerulli Associates, 2011. The Cerulli Edge: U.S. Asset Management. Research report, January.

Chen,Jun,SusanHudson-Wilson,andHansNordby,2004. Real Estate Pricing: Spreads & Sensibilities: Why Real Estate Pricing Is Rational. Journal of Real Estate Portfolio Management 10(1):1–21.

Chen, Lijian, and Thomas Mills, 2004. Global Real Estate Investment Going Mainstream.UBSRealEstate Research.

Conner, Philip, and Youguo Liang, 2005. Income and Cap Rate Effects on Property Appreciation: Some Guidance for Future Pricing Trends. The Journal of Portfolio Management,SpecialIssue,pp.70–79.

Craft, T.M., 2001. The Role of Private and Public Real Estate in Pension Plan Portfolio Allocation Choices. Journal of Real Estate Portfolio Management7(1):17–23.

Fisher, Jeff, David Geltner, and Henry Pollakowski, 2006. A Quarterly Transactions-Based Index of Institutional Real Estate Investment Performance and Movements in Supply and Demand. MIT Center for Real Estate. Available at web.mit.edu/cre/research/credl/pdf/credl-wp-0601.pdf

Kaiser, Ronald W., 2002. Public REITs vs. Private Real Estate: Assessing Your Options.Bailard,Inc.

Kapas, Manidipa, and Youguo Liang, 2010. A Bird’s Eye View of Global Real Estate Markets: 2010 Update. Prudential Real Estate Investors.

Lee, Stephen, and Simon Stevenson, 2005. The CaseforREITsintheMixed-AssetPortfoliointheShort-andLong-Run.Journal of Real Estate Portfolio Management11(1):55–80.

Lindahl, David P. Making an Allocation to Real Estate. Kennedy Associates Real Estate Counsel, Inc.

Ling, David C., Andy Naranjo, and M. Nimalendran, 2000. Estimating Returns on Commercial Real Estate:ANewMethodologyUsingLatent-VariableModels. Real Estate Economics 28(2):205–231.

Lizierie,Colin,andCharlesWard,2000.Commercial Real Estate Return Distributions: A Review of Literature and Empirical Evidence. University of Reading Working Paper No. 01/00.

Mueller, A.G., and G.R. Mueller, 2003. Public and PrivateRealEstateinaMixed-AssetPortfolio.Journal of Real Estate Portfolio Management 9(3):193–203.

Pagliari, Joseph L. Jr., and James R. Webb, 1995. A FundamentalExaminationofSecuritizedandUnsecuritizedRealEstate.The Journal of Real Estate Research 10(4):381–426.

Pagliari, Joseph L. Jr., Kevin A. Scherer, and Richard T. Monopoli, 2003. Public Versus Private Real Estate Equities:ARisk-ReturnComparison.The Journal of Portfolio Management,SpecialEdition,pp.101–111.

Philips,ChristopherB.,2011.Considerations for International Equity. Valley Forge, Pa.: The Vanguard Group.

Seiler, Michael J., James R. Webb, and F.C. Neil Myer,2001.CanPrivateRealEstatePortfoliosBeRebalanced/Diversified Using Equity REIT Shares? Journal of Real Estate Portfolio Management 7(1):25–41.

P.O.Box2600 ValleyForge,PA19482-2600

Connect with Vanguard® > vanguard.com

© 2011 The Vanguard Group, Inc. All rights reserved. Vanguard Marketing Corporation, Distributor.

ICREECR 032011

Vanguard research >Vanguard Center for Retirement Research Vanguard Investment Counseling & Research Vanguard Investment Strategy Group

E-mail > research@vanguard.com

ETFs are not redeemable with an Applicant Fund other than in Creation Unit aggregations. Instead, investors must buy or sell ETFs in the secondary market with the assistance of a stockbroker. In doing so, the investor will incur brokerage commissions and may pay more than net asset value when buying and receive less than net asset value when selling.

For more information about Vanguard funds, visit vanguard.com, or call 800-662-2739, to obtain a prospectus. Investment objectives, risks, charges, expenses, and other important information about a fund are contained in the prospectus; read and consider it carefully before investing

CFA® is a trademark owned by CFA Institute.