Analysis of REIT IPOs Using a Market Microstructure Approach: Anomalous Behavior or Asset Structure

In this research, we examine and present new evidence on the market activity following the initial public offering (IPO) of a real estate investment trust (REIT) using microstructure data. We analyze the bid-ask spread differences for REIT securities compared to common stocks and closed-end funds for all IPOs between 1985 and 1988. Our results show that REITs, as a whole sample, experience significantly greater bid-ask spreads immediately following the IPO compared to common stocks and funds. However, this outcome is driven by the equity REIT sample, with the mortgage REIT sample having significantly smaller bid-ask spreads. This is in contrast to the evidence documented by Nelling, Mahoney, Hildebrand, and Goldstein (1995). We attribute our result to the underlying asset structure (such as equity, hybrid, and mortgage portfolios) of the various REITs. Overall, however, we find that bid-ask spreads for REITs are similar to those of common stock when both asset structure and the traditional determinants of the spread (share price, trade volume, and returns variance) are considered.     

How Much Do REITs Pay for Their IPOs?

Previous studies have documented the reversal in the initial returns of REIT IPOs from overpricing in the 1980s to underpricing in the 1990s. We find that the gross spreads of REIT IPOs decreased significantly in the 1990s. In particular, there is a bimodal clustering for gross spreads at 6.5 and 7.0%. Moreover, in the 1980s around 94% of REIT IPOs had integer offer prices, most of which were priced at either $10 or $20. However, the proportion of integer offer prices decreased to 64% in the 1990s. Higher gross spreads, overpricing, and high frequency of integer offer prices for REIT IPOs in the 1980s are consistent with the marketing hypothesis that in the 1980s REIT IPOs were mainly marketed to less-informed individual investors. Our results explain the dynamic process employed by underwriters in the setting of gross spreads and the pricing of REIT IPOs as a new financial product in response to various structural changes in REITs.     

An Analysis of Lockups in REIT IPOs

We analyze how the unique characteristics of real estate investment trusts (REITs) affect IPO lockup agreements from 1980 to 2006. The findings show that, unlike industrial IPOs, lockup periods for REIT IPOs do not cluster at 180 days, tend to cover longer periods, and vary over time. Our results support the commitment device hypothesis instead of the signaling hypothesis. That is, REIT managers tend to use lockup agreements to alleviate moral hazard problems and protect post-IPO investors rather than to send signals to investors. Finally, contrary to previous studies, we find no significant negative abnormal returns around the unlock date for the whole sample. The lack of aggressive sales by insiders and the fact that REITs are not backed by venture capitalists can explain our finding.     

REIT Stock Splits and Market Efficiency

An analysis of real estate investment trust (REIT) stock splits is presented. Evaluation of the initial reaction to split REITs supports efficient market pricing where REITs generate statistically significant positive announcement date returns, no statistically significant record date returns, and muted ex-date returns. In the long-term, split REITs do not consistently out perform benchmark portfolios over one-year, two-year, and three-year periods. REITs split subsequent to a substantial run up in stock price and to improve the position of their post split stock price relative to the stock price of the typical REIT.     

Are REIT IPOs Unique? The Global Evidence

Previous studies of real estate investment trust (REIT) IPOs have focused primarily on REITs listed in the U.S. These studies in general find that, unlike industrial firm IPOs, REIT IPOs in the U.S. exhibit an abnormally low initial-day return and mixed long-run performance. Our study examines this puzzle using a large sample of 370 REIT IPOs from four continents (14 different countries) during the 1996–2010 period. We find that (1) the newly-established REITs in other countries exhibit similar initial-day return pattern as in the U.S., (2) the low initial-day return might be caused by the fund-like structure of REITs and the re-deployable assets (real estate) they hold, (3) the slightly positive initial-day return is offset by the poor performance in the 190 days subsequent to the IPO, and (4) the change in U.S. REIT IPO performance before and after 1990 is likely due to a change in the REIT structure.     

JANUARY RETURN SEASONALITY IN REAL ESTATE INVESTMENT TRUSTS: INFORMATION VS. TAX-LOSS SELLING EFFECTS

We examine the January return seasonality of real estate investment trust (REIT) common stock and underlying assets. Both stock returns and the National Assocation of Realtors median home price index exhibit January seaonals. However, the median home price index explains little of the seasonal stock returns, and a significant January effect in stock returns remains for small REITs. Thus, information effects are not the likely cause of the January effect in REITs. Further analysis indicates that tax-loss selling is the more likely cause of the January effect.     

Investor Sentiment and REIT Returns

Recent advances in the field of behavioral finance have given a fillip to the use of behavioral factors in asset pricing models. This study adds to the understanding of the REIT return generating process by exploring the behavioral impact of investor sentiment on REIT returns. The results show that when investors are optimistic (pessimistic), REIT returns become higher (lower). These findings are robust when conventional control variables are considered. Empirical analysis indicates steady erosion in the importance of the default and term structure interest rate variables previously considered as important determinants of REIT returns. Previous noise trading papers that consider the impact of institutional traders conclude that institutional investors cannot arbitrage away noise trader risk. The results of this paper find an exception in the case of small REITs. Examination of REITs based on size reveals that the return generating process of small REITs differs from that of mid-size and large REITs. Analysis of the return generating process by performance shows high performance REITs are more sensitive to the independent variables in the model as compared to the low and mid performance REITs.     

REIT Short Sales and Return Predictability

We examine REIT short sale transactions and show REITs are shorted less frequently than non-REITs. Results also show short sellers are less able to predict negative future returns for REITs, relative to non-REITs, which is consistent with increased pricing efficiency for REITs and suggests REIT assets are more transparent. In a broader context, these results suggest differences in transparency across asset types influence the effectiveness of short selling. Results showing REIT short sellers are contrarian imply traders target REITs that are performing well instead of underperforming REITs, suggesting restrictions on REIT short sales should be re-evaluated.     

REIT Characteristics and the Sensitivity of REIT Returns

Previous research on the returns to real estate investment trusts (REITs) has considered whether REITs are systematically exposed to general stock-market risk and interest-rate risk. This study examines how the sensitivity of REIT returns to these factors may be influenced by various REIT characteristics. Using a sample of publicly traded REITs, we estimate the sensitivity of REIT returns to stock market and interest-rate changes. We then propose and implement a model for testing whether differences in asset structure, financial leverage, management strategy, and degree of specialization in the REIT portfolios are related to their sensitivity to interest rate and market risk. Our results permit us to offer some inferences about how REITs can alter their risk exposure by managing these characteristics.     

The Relative Importance of Stock, Bond and Real Estate Factors in Explaining REIT Returns

This paper examines the link between REIT, financial asset and real estate returns, and tests whether it changed subsequent to the “REIT boom” of the early 1990s. The main focus is on answering the question do REIT returns now better reflect the performance of underlying direct (unsecuritized) real estate? We develop and implement a variance decomposition for REIT returns that separates REIT return variability into components directly related to major stock, bond, and real estate-related return indices, as well as idiosyncratic or sector-specific effects. This is applied to aggregate REIT sector (NAREIT) returns as well as returns to size and property-type based REIT portfolios. Our results show that the REIT market went from being driven largely by the same economic factors that drive large cap stocks through the 1970s and 1980s to being more strongly related to both small cap stock and real estate-related factors in the 1990s. There is also a steady increase over time in the proportion of volatility not accounted for by stock, bond or real estate related factors. We also find that small cap REITs are “more like real estate” compared to larger cap REITs, at least over the 1993–1998 period. We argue that this could be a result of the institutionalization of the ownership of larger cap REITs that took place in the 1990s.     

REIT IPOs and the Cost of Going Public

We examine Initial Public Offerings (IPOs) of Real Estate Investment Trusts (REITs) that went public between 1986 and 2004. Consistent with previous studies, we find that REIT IPOs are associated with lower levels of underpricing relative to traditional issues. We also find that REITs are associated with smaller file price revisions. Both findings are potentially attributable to the lower level of uncertainty associated with pricing REITs. In contrast, using an alternative measure of issuance costs that incorporates the share retention decision by preexisting owners, we find no significant difference between REIT and non-REIT issues, suggesting the results of previous studies are not robust to various specifications of issuance cost and that preexisting owners do not necessarily benefit from the lower level of underpricing. Additionally, we find no difference in the issuance costs of equity versus mortgage REITs, particularly once we control for the use of umbrella partnerships.

On the Hybrid Nature of REITs

The consensus that emerges from the current research on the linkage between securitized and direct investment in real estate is that direct (private) real estate returns play a relatively minor role in the real estate investment trust (REIT) return generating process. However, this result may at least partially be due to the coarseness of the measures of direct real estate returns or the relatively short return horizons used in previous studies. This study takes a different and unique perspective. Unlike earlier studies we do not use aggregated, average appraisal based returns on direct real estate investment. Instead, we use the MIT TBI indexes, which are transaction based price indexes, available both on the aggregate and sub-index levels. We find that the relation between REIT and direct real estate returns appears to be stronger at longer horizons. More specifically, using a cointegration framework, we find robust evidence that REITs and the underlying real estate are related and that they share a long run equilibrium. Interestingly, we find that both REITs and direct real estate returns adjust towards this long run relationship. When we examine property type level data we find similar results.     

The Cost Efficiency of Real Estate Investment Trusts: An Analysis with a Bayesian Stochastic Frontier Model

Using a stochastic frontier methodology that incorporates Bayesian statistics, this paper analyzes the cost efficiency of real estate investment trusts (REITs) by observing the deviations of the measured costs of individual REITs from a defined efficient cost frontier. Using 1995–1997 data, we extend the previous research in this area and measure REIT efficiency more precisely by isolating random measurement error from the overall deviations from the efficient cost frontier. We calculate the magnitude of each REIT's managerial inefficiency, the industry inefficiency, and returns to scale. In addition, we assess specific characteristics of REITs for their contribution to inefficiency by calculating the odds ratio that a REIT with a specific characteristic is more efficient than a REIT with an alternative characteristic. The results show that, for the years studied, REITs are relatively cost efficient with most REITs facing increasing returns to scale. Additionally, the REIT's use of debt and the REIT's management style significantly affect the cost performance of REITs during the aforementioned time period. Finally, diversification across property types, as measured, does not seem to influence REIT cost efficiency.     

Tracking the Evolution of Idiosyncratic Risk and Cross-Sectional Expected Returns for US REITs

This paper adopts the methodology in Bali and Cakici (Journal of Financial & Quantitative Analysis, 43, 29–58, 2008) in tracking the evolution of the relation between equity REITs’ idiosyncratic risk and their cross-sectional expected returns between 1981 and 2010. In addition to the full sample period, we study this relation for (i) January 1981–December 1992, (ii) January 1993–September 2001, (iii) November 2001–August 2008 and (iv) November 2001–December 2010 and produce empirical results for (i) all sample REITs, (ii) REITs with a price greater than $10 or (iii) REITs with a price greater than $5. Each period represents different dynamics (including the Global Financial Crisis) in the life of the REIT industry and leads to a different hypothesis. Further, we present comparative results based on the Fama-French 3- and 4-factor models. Overall, we document a negative relation between idiosyncratic risk and cross-sectional expected returns and demonstrate that this negative relation changes over time. These findings amplify the “idiosyncratic volatility puzzle,” as reported in the recent finance literature. Interestingly, REITs with a price of $5-to-$10 do well in 2009 and 2010. Further, the momentum factor appears to be influential since the first-ever listing of a REIT in the S&P500 Index in early October 2001.     

What Determines Stock Price Synchronicity in REITs?

This paper studies the behavior of REIT stock price synchronicity for the years 1997 through 2006. Theory suggests that REIT stock prices should be largely independent of market changes; and, at the very least, REITs should have a low covariance with other assets, including other REIT stocks. The evidence presented below does not support this view. Instead, synchronicity appears to be quite high in the equity REIT market, especially among REITs that larger and more liquid. We also find that REIT stock price synchronicity is negatively related to hedge fund ownership, but positively related to pension fund and insurance company ownership. The evidence further suggests that synchronicity is the highest among industrial and regional mall REITs, and lower among apartment, health care, and mixed property REITs.     

Equity Capital Flows and Demand for REITs

REITs are attractive to investors due to their unique characteristics such as high dividend yields, low correlation with common stocks, and a potential hedge against inflation. Thus the market demand curve of REIT equities may not be horizontal. This paper examines the shape of the market demand curve for REIT equities by employing REIT equity capital flows as a proxy for REIT aggregate demand. Our results do not support a downward demand curve for REIT equities. That is, we do not find evidence for the price-pressure effect in REIT returns. Instead, we find it is REIT returns that affect REIT equity capital flows rather than REIT equity flows that affect REIT returns. The results are consistent when we allow for the presence of market fundamental variables in our analysis. In addition, a variance decomposition analysis suggests that REIT equity capital flows do not cause revisions in expected cash flows (dividends) that are strong enough to impact REIT returns. Thus our findings are consistent with implications that the market demand curve for REIT equities is horizontal.     

The Asymmetric Response of Equity REIT Returns to Inflation

This paper posits that the failure of past studies to document a positive relationship between REIT (Real Estate Investment Trust) returns and inflation is an artifact of the empirical framework that has predominated in these studies. Applying a pooled estimation methodology to an expansive data set containing 195 publicly traded equity REITs for the period 1981–2002, the study documents a strong asymmetry in the response of equity REIT returns to inflation. Specifically, when expected and unexpected inflation are separated into positive and negative changes, results indicate that equity REIT returns rise in response to both increases and decreases in inflation. The evidence, which is partly contingent on the prevailing monetary policy environment, carries important policy implications for portfolio management and provides insights into the observed anomalous relationship between REITs and inflation.     

Asymmetric Correlation and Volatility Dynamics among Stock, Bond, and Securitized Real Estate Markets

We apply a multivariate asymmetric generalized dynamic conditional correlation GARCH model to daily index returns of S&P500, US corporate bonds, and their real estate counterparts (REITs and CMBS) from 1999 to 2008. We document, for the first time, evidence for asymmetric volatilities and correlations in CMBS and REITs. Due to their high levels of leverage, REIT returns exhibit stronger asymmetric volatilities. Also, both REIT and stock returns show strong evidence of asymmetries in their conditional correlation, suggesting reduced hedging potential of REITs against the stock market downturn during the sample period. There is also evidence that corporate bonds and CMBS may provide diversification benefits for stocks and REITs. Furthermore, we demonstrate that default spread and stock market volatility play a significant role in driving dynamics of these conditional correlations and that there is a significant structural break in the correlations caused by the recent financial crisis.     

Depreciation-Related Capital Gains,Differential Tax Rates,and the Market Value of Real Estate Investment Trusts

We develop a model for valuing U.S. real estate investment trusts (REITs) that considers the tax liability impounded in REITs’ property portfolios. This liability is a function of the portfolio’s accumulated depreciation and is driven by different tax rates applied to individual components of the total gain from property sales. These two components are the capital gain resulting from the sale of property at a price higher than its cost and the gain due to the recapture of depreciation taken during the use of the property. Our measure of value is the REIT’s net asset liquidation value (NALV). The metric of REIT value currently used by analysts is a REIT’s net asset value (NAV), but a REIT’s NAV will always be greater than the NALV and therefore overestimate market value, all else equal. Finally, using observed market prices for REITs, we provide evidence that NALVs give superior estimates of REIT market prices than do NAVs.     

Returns transmission, value at risk, and diversification benefits in international REITs: evidence from the financial crisis

We examine daily cross-market return interactions and downside risk between a US REIT returns index and the return indexes of twelve international REIT markets. These relationships are investigated for a period of normal REIT market conditions as well as for periods of inflating and collapsing REIT prices. We find that US REIT returns are contemporaneously correlated with other REITs most strongly during the bubble and crash market conditions where the US REIT market is an almost unilateral transmitter of returns. We also find that the Value at Risk (VaR) of the least capitalized REIT markets is proportionally higher during base/normal market conditions but that the largest REIT markets have the highest VaR contribution during the crash (financial crisis) period. Overall, our evidence indicates that REIT market risk shifted to the largest REIT markets and that diversification benefits eroded considerably during turbulent market conditions.