Real Estate Mutual Funds Shopping Malls or Self-Storage?

Transcription

1 Real Estate Mutual Funds Shopping Malls or Self-Storage? Steve P. Fraser and H. Shelton Weeks Florida Gulf Coast University Abstract Investors recognize the importance of asset allocation. However, one must clearly understand the characteristics of any asset class added to the portfolio. Here we examine the holdings of real estate mutual funds to examine the nature and level of property-type diversification. We find that real estate funds hold a wide array of property types in varying degrees of concentration. These factors can dramatically affect portfolio composition. The results of this research suggest investors need to exercise extreme caution when selecting a real estate mutual fund to add to their financial portfolio. JEL Classifications: G11 Keywords: asset allocation, REIT, mutual fund

2 1. Introduction Investment professionals often coach individual investors to diversify their investment portfolio. Asset allocation, experts purport, is the critical driver of a portfolio s total return. Ibbotson and Kaplan (2000) suggest more than 90 percent of the variation in a portfolio s return over time is due to its strategic asset allocation. Similarly, Brown and Tiu (2007) find the strategic asset allocation decision of endowment funds explains more than 70 percent of the variation of those funds returns. An essential aspect of the allocation decision is the determination of how many, and which, asset classes to include in the portfolio. Many professional-managed portfolios, such as target-date retirement funds, often hold a simple portfolio containing allocations to equity, fixed income, and perhaps cash. As these retirement plan vehicles have evolved, Maxey (2007) notes the trend for funds to further diversify by adding additional asset classes specifically real estate. The move to expand the number of asset classes in the individual retirement portfolio space follows the portfolio allocation examples exhibited by institutional investors. For example, The National Association of College and University Business Officers (NACUBO) 2008 survey of colleges and universities found endowments with more than $500 million had allocations to real estate of six percent or greater. The Investment Fund for Foundations (TIFF), an investment advisor available for foundations and endowments, has a three percent allocation to real estate (specifically Real Estate Investment Trusts REITs) in its Multi-Asset Fund. Individual investors are not immune for the call to real estate as well. Malkiel (2003) suggests investors should not only seek to own their own home, but also perhaps own commercial real estate through REITs. For the individual investor, it is worth a brief divergence into the various viewpoints of real estate. Many individuals consider an investment in real estate simply as the investment in their home. Even with this simple perspective on real estate, there is considerable debate on the role and impact of the individual residential home in the portfolio. With a broader perspective on the asset class, Hudson-Wilson (2000) outlines what she describes as the modern structure of real estate investment options for institutional investors: private commercial equity, private commercial debt, public equity (e.g. RIETs), and public commercial debt (commercial mortgage-backed securities (CMBs)). An investment in one s home might be the equivalent private equity play for individual investors, and with access to REITs, investors can access a greater segment of the real estate asset class including CMBs. In short, with the possible exception of the private debt quadrant, the remaining investment objectives are available to individuals and institutions alike. Individuals can decide whether to own their own home, and then move to compliment the real estate allocation with further investment in the asset class. The securitization of today s real estate market provides investors with many options, including those with either a debt or equity bias if desired. The primary purpose of this study is to examine one of the real estate investment choices available to individual investors. If an investor wishes to invest in securitized real estate, she has two primary options. The first choice is to invest directly in a REIT. The purchase of a REIT will provide exposure to a diversified portfolio of properties; however in most cases; the properties will all be in a very narrow sector of the real estate asset class (e.g. regional malls, shopping centers, or even more specialized categories like timber). The second choice, and the 1

3 focus of this study, is for an investor to purchase shares in a real estate mutual fund which interestingly, most often hold REITs. The purchase of a real estate mutual fund might likely result in a more diverse array of properties held by the investor, yet the extent of the exposure to a given sector or the level of diversification is unknown without specifically examining the holdings of the specific mutual fund. In contrast to many equity mutual funds, where the name of the fund provides some insight into the objective (e.g. small value or large growth), real estate mutual funds most often carry the simple real estate fund moniker. We seek to investigate the composition of real estate mutual funds. We examine both the individual holdings, and the level of those holdings, contained in each fund. The results of this research are important for investors and practitioners alike. Investors addressing the need for further diversification must understand the degree to which investing in a real estate mutual fund in fact diversifies the portfolio. For planners and advisors, a better understanding of the characteristics of real estate funds will better assist their ability to serve their clients. This research builds on similar research in the area of target-date retirement mutual funds. Today, more and more individual investors are turning to professionally managed target-date funds for retirement. These one-stop funds allow investors to simply select a target retirement date, and let the finance professionals manage the asset allocation over time. Critical to the management of these funds is how the fund manages risk most often by adjusting the equity exposure of the fund (downward) as time approaches the specified target date. Nagengast, Bucci, and Coaker (2006) illustrate there are significant differences across fund sponsors in their approach to managing the equity exposure. For example, the percentage allocated to equity for a target date fund with a specific retirement date can vary greatly among the different mutual fund sponsors. Similarly, we examine the nuanced differences of property type allocations to investigate how the holdings differ across real estate mutual funds. The results provide insight into the degree of diversification (or lack of) provided by real estate mutual funds. There is limited research that investigates the holdings and sector diversification within real estate mutual funds. Previous research in the area focused primarily on examining the return performance of real estate mutual funds. O Neal and Page (2000) suggest that a sample of 28 real estate mutual funds does not provide any abnormal return. Gallo, Lockwood, and Rutherford (2000) find some evidence of abnormal return in real estate mutual funds and suggest managers add value to real estate portfolios by their investment decisions made across property types not with specific properties within a category or property type. They suggest any abnormal returns are due to overweighting outperforming property types. This conclusion supports the Ibbotson and Kaplan (2000) suggestion that the allocation decision remains of vital importance to the returns of a portfolio. So what does the allocation landscape of real estate mutual funds suggest? 2. Methodology We examine the composition of real estate mutual funds by developing a dataset of real estate mutual funds specific holdings. We use Standard & Poor s NetAdvantage to identify an initial set of real estate mutual funds resulting in 256 funds. These funds, however, are not necessarily unique holdings. A majority of real estate funds are offered as part of multiple share class (MSC) funds, and therefore many of the funds are simply sub-classes of the same portfolio of 2

4 holdings. Similar to O Neal and Page (2000), we select only one class from those funds offering multiple share classes, resulting in 74 funds. NetAdvantage identifies the top ten holdings of each fund as well as the percentage that each holding represents of the fund s total assets. When we examine the top ten holdings across the sample, we find 152 individual holdings. Virtually all of the holdings are individual REITs. We use the SNL Financial database to obtain the industry focus (regional mall, shopping center, etc.) for each REIT in their database. We then replace each mutual fund holding with the property type reported by SNL allowing us to better capture the nature of the property type allocation. We eliminate those funds where a majority of the top holdings could not be classified or if the majority of top holdings were not RIETs (e.g. repurchase agreements, cash equivalents etc.). The resultant sample contains 63 funds. 3. Results The research of Gallo, Lockwood, and Rutherford (2000) suggests that different property type categories provide different returns over time. This is no different than suggesting that value stocks and growth stocks provide different risk and return characteristics. In our attempt to extend their work, we first examine the performance of property types held by real estate mutual funds. We obtain monthly property focus index values from SNL Financial over the period and compute annual holding period returns. Panel A of Table 1 depicts the geometric return and the standard deviation of each series. Geometric returns range from a low of two percent (hotels) to 15 percent (regional malls and industrial). Those with the highest returns are not necessarily associated with the highest risk. Standard deviations range from a low of 15 percent (industrial) to a high of 30 percent (hotels). The ranges of risk and return characteristics highlight the importance of sector selection by a real estate mutual fund manager. Figure 1 plots the Table 1 data in risk and return space. Panel B of Table 1 presents the returnto-risk the ratios for the various sectors. Given the significant weights these mutual funds hold in individual REITs, forming a portfolio of REITs across different sectors can yield remarkably different results. When assets are combined, the correlation among the assets is a key driver of the risk profile of the portfolio. Panel C of Table 1 shows the correlation matrix for the SNL focus indices. We see a wide range of correlations among sectors highlighting the potential impact on the portfolio composition decision. We next investigate the individual holdings of the real estate mutual funds focusing on the degree of diversification found among funds. Table 2 reports the percentage of the fund s assets invested in the largest individual holding, invested within the most heavily weighted sector, and the aggregate investment represented by the fund s Top 10 holdings. The mean percentage for the funds largest individual holding is 8.37 percent. Recall that Gallo et al (2000) suggest real estate returns are dependent upon property types. Hence, we should expect that the returns of real estate mutual funds are significantly impacted by these large positions. Real estate mutual fund holdings appear to be considerably more concentrated than the holdings of large stock equity mutual funds. For example, we find the mean for Top 10 holdings within a fund to be nearly 50 percent of fund assets. Further, the percentage of assets invested in the Top 10 holdings is more than 40 percent of the portfolio value for 88 percent of the funds. 3

5 Table 1 Focus Index Returns, Risk & Correlations Panel A depicts the geometric means and standard deviations for the annual returns for each SNL Focus Index. Panel B depicts the return-to-risk ratio. Panel C depicts the correlations among sectors.. Diversified Healthcare Hotel Industrial Multifamily Office RegMall ShopCtr SelfStor Panel A Geo Mean Std Dev Panel B Return/Risk Panel C Diversified 1.00 Healthcare Hotel Industrial Multifamily Office RegMall ShopCtr SelfStor Figure 1 Return SNL Focus Sectors Industrial ShopCtr Diversified RegMall SelfStor Mul5family Office Healthcare Hotel Standard Devia.on Table 2 illustrates the nature of the holdings in the sample. In contrast, in American Funds The Growth Fund of America, one of the largest equity mutual funds, the Top 10 holdings comprise 4

6 less than 20 percent of the fund s total assets. Extending our analysis beyond the top holding, we also examine the investment in the largest sector. We find the mean percentage investment in the most heavily weighted sector is more than 13 percent of fund assets. Table 2 Holdings Snapshot Examining the Top 10 holdings of each fund, this table reports the percentage of fund assets invested in the fund s largest single holding; invested within the single largest sector, and the percentage of total fund assets invested in the Top 10 holdings Mean Std Dev Min Max Largest Holding Largest Sector Total of 10 largest N=63 A fund manager s selection of property types (and in many cases, the individual REIT) can have significant implications for the portfolio because so much of the fund s assets are invested in the Top 10 holdings. The most common individual holding appearing in funds in the sample is Simon Property Group (regional mall), included in 57 of the 63 funds. In fact, Simon is the largest holding in 51 funds. Vorando (diversified) and Boston Properties (Office) each appear in 50 funds, although only Boston Properties appears as a top holding, and only once. Concentration of holdings amplifies the risk of each holding. Consider the impact of a single poor-performing REIT within a fund s top holdings. General Growth Properties (regional mall), which fell more than 80 percent in value in 2008, appears in 27 real estate mutual funds, although only once as the top holding. In an effort to further characterize the composition of real estate mutual funds, we compute the mean investment in each property sector and the concentration that these sectors represent across the Top 10 holdings. Table 3 reports the mean investment in each sector. The regional mall sector is by far the dominant sector represented in the sample, representing more than 10 percent of fund holdings. While the Healthcare and Specialty sectors appear to have the lowest mean investment by funds in the sample, the results can be misleading. In an effort to better understand the level of diversification across sectors, we compute a concentration ratio, defined as the percentage of assets held in the largest sector divided by the percentage represented by the Top 10 holdings. A larger ratio suggests a lower level of diversification across property sectors, and therefore increased risk for the fund. A lower concentration ratio suggests a greater degree of diversification, within the Top 10 holdings. We sort the funds by level of diversification. Table 4 reports the five funds with the highest concentration ratio (lower diversification) and lowest concentration ratios (higher diversification). In sharp contrast to the results suggested in Table 3, Table 4 suggests that some of the more concentrated funds are heavily weighted in the health care and specialty sectors as opposed to the more prevalent regional mall sector. 5

7 Table 3 Focus Holdings Examining the Top 10 holdings of each fund, this table reports the mean percentages of fund assets invested in different property sectors. Variable Mean Std. Dev. Min Max Regional Mall Office Multi-Family Industrial Shopping Center Diversified Self Storage Hotel Health Care Specialty N=63 Table 4 Fund Concentrations This table reports the highest and lowest concentration ratios for real estate mutual funds. This concentration measure is the ratio of the percentage of fund assets invested in the largest propertytype sector divided by the percentage of fund assets invested in the Top 10 holdings. Largest Largest Top10 Ratio Sector Wt Wt Multifamily Multifamily RegMall RegMall RegMall RegMall Healthcare Diversified Healthcare Specialty

8 The results of this study are subject to limitations and cautions for investors. Using a holdingsbased approach, the data utilized here are subject to the practice of window dressing where managers change portfolio holdings just prior to the end of reporting periods. Further, our examination is limited to the conclusions that can be based upon examination of the largest ten holdings of each fund. This limitation notwithstanding, we find the large percentage of assets invested in only ten securities (up to a maximum of 68 percent), not only particularly telling, but also especially relevant to the portfolio composition decision for investors and advisors. 4. Conclusion Individual investors and financial planners face choices with respect to how they might add real estate to an investment portfolio. One might invest in a REIT, or in a mutual fund that invests in REITs. In this study we examined the holdings of real estate mutual funds to provide greater insight into the sector diversification across funds. The results of our examination suggest there is a wide disparity of exposure to individual real estate sectors among funds that can greatly affect their contribution to a portfolio. As a result, the impact on a portfolio of adding a real estate mutual fund is highly dependent on the fund selected for inclusion. While real estate is often associated with the tag line: location, location, location, the same might be true for selecting a real estate mutual fund from the spectrum of choices available. The location of a fund within this spectrum will have a significant impact on the fund s contribution to the overall portfolio. Hence, investors and planners must exercise caution when seeking to diversify by adding real estate mutual funds to their portfolios. 7

9 References Brown, K., Garlappi, L., and Tiu, C., The Trove of Academe: Asset Allocation, Risk Budgeting, and the Investment Performance of University Endowment Funds, working paper, Gallo, J., Lockwood, L. and Rutherford, R., Asset Allocation and the Performance of Real Estate Mutual Funds, Real Estate Economics, 2000, 28(1), Hudson-Wilson, S., Modern Real Estate Portfolio Management, 2000, New Hope PA: Frank J. Fabozzi Associates. Ibbotson, R. and Kaplan, P., Does Asset Allocation Policy Explain 40, 90, or 100 Percent of Performance? Financial Analysts Journal, 2000, 56(1), Malkiel, B., A Random Walk Down Wall Street, 2003, New York NY: W.W. Norton & Company. Maxey, D., Target-Date Funds Spread Wings; Firms Add REITs, High-Yield Bonds as Field Evolves, The Wall Street Journal, 2007, May 10, D6. NACUBO, National Association of College and University Business Officer Endowment Survey, 2008, Last accessed March Nagengast, J., Bucci, J. and Coaker, W., Popping the Hood An Analysis of Major Life Cycle Fund Families, Marina del Rey CA, Turnstone Advisory Group LLC, O Neal, E. and Page, D., Real estate mutual funds: Abnormal Performance and Fund Characteristics, Journal of Real Estate Portfolio Management, 2000, 6(3), TIFF, The Investment Fund for Foundations, last accessed July