Real Estate Within the Asset Allocation Mix

Table of Contents

List of Abberviations

List of Figures and Tables

1 Introduction
1.1 Problem Definition and Objectives
1.2 Course of the Investigation

2 Theoretical Fundamentals
2.1 Portfolio Fundamentals
2.2 Definition of Real Estate

3 Real Estate Investment Discussion
3.1 Risk and Return
3.1.1 Direct Real Estate
3.1.2 Real Estate Investment Trusts (REITs)
3.2 Diversification in a Mixed-Asset Portfolio
3.2.1 Direct Real Estate
3.2.2 Real Estate Investment Trusts (REITs)
3.3 Real Estate as Inflation Hedge
3.4 Disadvantages of Real Estate Investment

4 Optimal Allocation

5 Summary and Conclusion

Reference List

Appendix

List of Abbreviations

illustration not visible in this excerpt

List of Figures and Tables

Figure 1: Efficiency Frontiers of Portfolios With and Without Real Estate (Appraisal Data Versus Unsmoothed Data)

Figure 2: Reasons for Investing in Real Estate

Table 1: Asset Correlation Matrix (1970-1995)

1 Introduction

1.1 Problem Definition and Objectives

Constructing smart portfolios is the key goal of every investor regardless of the risk aversion. Accessible investments for investors are for instance stocks, bonds, treasury bills, and real estate. According to Seiler, Webb, and Myer (1999, p. 163) “real estate asset management has been and will continue to be a topic of great interest”. In the year 1971 U.S. public real estate had a total market capitalization of US$1.4bn, while in 2006 public real estate had a market capitalization of US$438bn (National Association of Real Estate Investment Trusts [NAREIT], 2007, p. 1). The U.S. private real estate index has more than tripled from US$84bn in market value in the first quarter of 2001 to US$266m in the first quarter of 2007 (National Council of Real Estate Investment Fiduciaries [NCREIF], 2007, p. 1). It is obvious that the real estate market has been growing incredibly and real estate has became more and more important as an investment opportunity. However, all available data on ownership of real estate show that pension funds hold 3.5% to 4.0% of their total assets in real estate (Chiochetti, SA-AADU, & Shilling, 1999, p. 193). In contrast to that, Kallberg, Liu, and Greig (1996, p. 371) for example find that real estate should compose 9% of optimal mixed-asset portfolios, while others suggest even an allocation of 20% to 80% depending on the risk return preferences of investors and whether real estate is private or public (Benjamin, Sirmans, & Zietz, 2001, p. 197). Optimal allocation seems to be a problem. Another point is that some degree of diversification can be achieved without real estate. So why should investors hold real estate in their portfolios? Does real estate outperform stock and bond returns? What risks are linked with real estate investments? First of all, real estate is a very special class of assets. For example, the CAPM and APT assume a perfect arbitrage conditions under which any mispricing will be arbitraged away. The real estate market does not even approximate these conditions. For these types of assets non-risk characteristics such as marketability, transaction costs, and taxes must be taken into account and may be more important than risks in pricing (Ibbotson & Siegel, 1984, p. 220).

The aim of this paper is to provide the reader with a deep insight into the real estate investment discussion and to present the advantages and disadvantages of real estate in a mixed-asset portfolio. In a nutshell, at the end of this paper the reader should be able to decide, whether real estate investment is justifiable or not.

1.2 Course of the Investigation

This seminar paper is divided into three parts which are all based on each other. In the first step, the reader can find necessary theoretical fundamentals, which build the basis for the further handling of the topic. At this juncture first MPT and CAPM are presented and their relevance for real estate investment is explained. Second, a definition of private and public real estate is given with a focus on their abilities to generate return and the main sources of data, which are the basis for almost all studies in this paper, NAREIT and NCREIF indexes are described. Due to the limited coverage of this work, the focus is on U.S. real estate and international real estate as a further supplement. Of course single research results specific for a particular country can differ from results presented in this paper, but the tendency of results should be the same. In the main part based on the theoretical fundamentals advantages and disadvantages of both public and private real estate in a mixed-asset portfolio are presented. Concerning disadvantages, only private real estate is taken into consideration. In the fourth chapter possible optimal allocation of real estate in a mixed-asset portfolio is presented, differentiating between investors with different risk aversions. Finally, all results are summarized and the question is discussed whether real estate investment is justifiable.

2 Theoretical Fundamentals

2.1 Portfolio Fundamentals

Investors, also real estate investors, have two related goals when they build portfolios. First, they try to achieve a satisfactory overall return on portfolios they have to manage. Second, due to the fact that higher returns are generally obtained by accepting a higher risk, portfolio managers must ensure that they maximize the portfolio return for a given level of risk. The tools for achieving both goals are first selecting and second managing individual assets (Viezer, 1999). In his study Portfolio Selection H. Markowitz (1952) was the first who discussed the concept of diversification through the development of Modern Portfolio Theory (MPT). Although he used stocks to illustrate this, his theory can be applied to bonds, government treasury securities, other financial assets, and real estate as well. Markowitz showed that the total risk of a portfolio is a function not only of the assets’ means and standard deviations, but the degree to which their returns are correlated (p. 89). Correspondingly, it is theoretically possible for a portfolio manager to construct portfolios of zero risk by combining risky assets that are perfectly negatively correlated. In reality the risk free portfolio is unattainable because most assets exhibit some degree of correlation. Likewise, there is a limit to the benefits of diversification. The systematic risk or market risk, which can be attributed to macroeconomic factors, is always there and cannot be avoided through combining assets. In this context, real estate might warrant inclusion in a mixed-asset portfolio in terms of return contribution and lowering or removing unsystematic risk.

Thereafter W. Sharpe (1964), J. Lintner, J. Mossin, and J. Treynor independently developed the so-called Capital Asset Pricing Model (CAPM). The CAPM in particular takes a different perspective than the theory of Portfolio Selection. It does not deal with the issue how to construct a portfolio, instead it deals with the fact that all investors, who are more or less risk averse, hold a replica of the market portfolio. In other words, all more or less risk averse investors have identical expectations regarding risk and return of individual assets. The market portfolio is the most diversified portfolio comprising all assets in the same proportion as in the global market (Sharpe, 1964, p. 431). Considering this model, real estate should make up a certain part of the market portfolio and a part of every more or less risk averse investor portfolio.

2.2 Definition of Real Estate

Real estate is one of the world’s major types of assets. Similar to a single stock, real estate can both possess a great amount of risk and return benefit. First of all, it must be said that real estate can be divided into two main categories. One has to distinguish between private real estate or direct real estate and public real estate, the so-called Real Estate Investment Trusts, abbreviated REITs. National Association of Real Estate Investment Trusts collects and validates data regarding the performance of U.S. REITs in the so-called NAREIT index. U.S. direct real estate appraised data is collected and provided by the National Council for Real Estate Investment Fiduciaries in the so-called NCREIF Property Index (NPI).

Direct real estate is not publicly traded and is not divisible. One fundamental difference between direct real estate and all other asset classes is the way in which returns are calculated. Direct real estate indexes such as the NCREIF are based on appraisal data and not on transaction-based data (Seiler et al., 1999, p. 170). Appraisals can be done at any frequency, but most real estate properties are appraised once a year. For different properties appraisals are performed at different times throughout the year. The direct real estate indexes are annual indexes, which are updated quarterly (Seiler et al., 1999, p. 170). Direct real estate is linked with high transaction and information costs because it is not publicly traded. The marketability of direct real estate is low.

REITs are real estate companies with the purpose of purchasing and selling as well as managing properties. REIT must be (a) a corporation, business trust, or association;

(b) must invest 75% of its assets in real estate assets, cash items, and government securities;

(c) must generate at least 75% of gross income from real estate; and (d) must distribute 90% of taxable income to shareholders as dividends (Deutsche Börse AG, 2007, pp. 1-3). There are three different types of REITs that differ according to the business activity of the company. Equity REITs invest at least 75% of their total assets in income producing real estate properties. In other words, these companies rent out properties and generate their income in this way. The second group, mortgage REITs, invest at least 75% of their assets in residential mortgages and commercial mortgages and additionally grant loans for real estate projects (Peterson & Hsieh, 1997, p. 322). Finally, the third group are hybrid REITs, which are a mixture of the two. In 2007 equity REITs have been making up 90% of the total market capitalization of this sector (Lee & Stevenson, 2007, p. 541). In a nutshell, REITs can generate profit through income, for example rents, through additional services, for example real estate brokerage, and through disposal of properties. Like capital market securities REIT shares are traded on auction markets and can be acquired even by small investors. As a conclusion, REITs are divisible and have a great level of liquidity due to a high number of market participants.

3 Real Estate Investment Discussion

3.1 Risk and Return

3.1.1 Direct Real Estate

Adding a new asset class to the portfolio, an investor has to consider the contribution an asset makes to the risk and return performance of a portfolio as a whole and not separately its individual characteristics. Additionally, investments in real estate cannot be considered isolated, instead they must be put in context with other investment opportunities like stocks and bonds. The major criteria in the investment decision process are expected returns and risk. On the one hand in the studies presented below direct real estate based on appraisal data outperforms stocks and bonds concerning risk and return. On the other hand direct real estate based on transaction data does not provide better risk adjusted returns than other assets.

Hoag (1980, pp. 576-577) examined a sample of 463 U.S. industrial properties based on appraisal data over the period from 1973 to 1978. The result is that industrial direct real estate with an average return of 3.38% per quarter outperforms stocks and bonds. The standard deviation of 8.61% is high, but direct real estate still outperforms S&P 500 concerning the return to risk ratio. Giliberto (1990, p. 260), who examined U.S. annual real estate returns from 1978 to 1989, also states that direct real estate provides better risk adjusted returns than bonds and stocks with a mean of 11.31% and a standard deviation of 2.81%. However, the author indicates that these appraisal-based returns might dampen the measured volatility and overestimate the attractiveness of real estate (p. 260).

Ziobrowski and Ziobrowski (1997) compared U.S. direct real estate based on appraisal data and unsmoothed data over the period from 1970 to 1995 in their study. As expected, direct real estate based on appraisal data with an annual return of 8.43% and a standard deviation of 6.41% outperforms common stocks with a mean of 13.11% and a standard deviation of 16.35% on a risk adjusted level (p. 111). In contrast to that, the adjustment for smoothed data causes real estate returns to decrease and variance to rise. Even if the best estimate is regarded with an annual mean of 7.80% and 9.90% standard deviation, it clearly underperforms the above result and offers with a risk-return ratio of 0.8 return per unit of risk almost the same return performance as common stocks. Further evidence is provided by Webb, Miles, and Guilkey (1992) who compare appraisal data and transaction data of commercial direct real estate from 1980 to 1988. Appraisal-driven real estate returns with a quarterly mean of 2.5% and a variance of 5.0% understate the true variance of real estate properties of 3.1% mean and 6.9% variance (p. 351). The return to risk ratio also changes in the negative direction. Evidence on international level is provided by Hoesli, Lekander, and Witkiewisz (2004, pp. 171-174), who examined international direct real estate returns from 1978 to 2001. The conclusion of this study is that direct real estate exhibits in five of seven considered countries lower returns and higher standard deviations than bonds. At this point differences must be recognized between appraisal-based returns and true market returns, which are more directly comparable to returns for alternative investments. These differences drastically affect all statistics about the returns that are of interest to investors like the mean and the variance of an asset (Geltner, 1991, p. 327).

However, direct real estate also provides advantages when risk and return issues are considered. One advantage is that direct real estate returns are easy to forecast. Because appraisals occur infrequently changes in real estate values will be reflected more quickly by changes in REIT prices. With regard to that, REIT returns lead direct real estate returns, or in other words, appraisal based series incorporate market information with a lag (Seiler et al., 1999, p. 167). Gyourko and Keim (1992, p. 468) examined the lead-lag relationship of REITs and direct real estate and concluded that previous calendar year equity REIT returns are significant to explain current period NCREIF returns. Due to this predictability, portfolio weights can be altered to capture future unrealized returns.

Another advantage is that there is a real estate factor premium, which makes real estate in a mixed-asset portfolio indispensable. If the variables that drive returns for stock and bonds are the same as for real estate, then it is likely that real estate will not warrant its inclusion (Seiler et al., 1999, p. 164). Mei and Lee (1994, p. 121) investigated equity REITs and U.S. direct real estate and came to the conclusion that there is a real estate factor in addition to a bond and a stock factor in asset pricing. In general, real estate must be included in a portfolio to capture this factor premium.

[...]