|
Article Excerpt
A leading explanation for IPO cycles is time-varying supply and demand for the underlying assets of the firms that are considering going public. We test this hypothesis using REIT IPOs, taking advantage of the relative transparency of the underlying real asset markets. We document links between REIT IPO activity and both the conditions of the underlying real estate market and the price of REITs. We find no significant relation between the heat of the IPO market and post-IPO operating performance, implying homogeneous firm quality across IPO cycles. Finally, we show that lagged IPO proceeds are related to future increases in investment and in capacity utilization.
**********
Initial public offerings have provided a rich arena for theoretical and empirical analysis. The theoretical approaches are diverse and often center on the asymmetric information characteristics of IPOs: Given that managers of the firm know more about the firm's prospects than potential equity investors, how should markets react to an IPO announcement? On the one hand, we might expect managers to exploit their informational advantage and issue equity only when markets overvalue their assets. This would imply relatively poor long-term performance of IPOs, via a lemons or adverse-selection explanation. Conversely, high levels of IPO issuance might signal positive information about the industry's prospects or future investment opportunities. This demand for investment capital would lead firms to issue equity in order to undertake favorable projects. In addition, firms could also deliberately underprice in order to signal quality in anticipation of a subsequent seasoned equity offering, as in Welch (1989). These two possibilities would not predict poor future performance of IPO firms.
These conflicting theoretical predictions have spawned a very broad array of empirical research. Ibbotson and Ritter (1995) partition this literature into three parts: initial-day underpricing, long-term underperformance of IPOs and cycles of hot and cold IPO markets. While initial-day underpricing is almost universally found in the many domestic and international studies, the evidence on long-run behavior provides no clear consensus. Ritter (1991) computes an industry-adjusted, 3-year return of -15% and provides evidence that firms that issue in hot markets tend to be of poorer quality than their peers. In contrast, Helwege and Liang (2004) find little difference between firms that issue in a hot market and those that issue in a cold market.
Additional insight into these important questions can be obtained by focusing the analysis on a restricted class of equities, real estate investment trusts (REITs), rather than on all IPOs. This study empirically addresses these issues by examining the relation between IPO performance and the underlying real assets. We study 189 REIT IPOs completed between the beginning of 1980 and the end of 1998. (1) We use REITs for this study for five major reasons: (i) to take advantage of the availability of detailed information on the underlying real asset markets; (ii) to focus on a single-industry sector (several theoretical papers on IPOs attribute the clustering of IPOs to positive shocks to an industry's prospects); (iii) to address further the conflicting literature on REIT IPOs; (iv) to analyze an industry where the problems of asymmetric information and managerial self-dealing may be mitigated due to transparency of the underlying real asset market and constraints on managerial discretion and (v) since REITs are a tax conduit, the decision to raise equity rather than debt is not influenced by tax effects.
Our study complements the earlier (sometimes conflicting) evidence on REIT IPOs. Wang, Chan and Gau (1992) document a period of REIT overpricing together with poor performance over the first 120 days of trading. Ling and Ryngaert (1997), using a later time frame, document underpricing for REIT IPOs and find positive abnormal performance up to 100 days after the offering. They attribute this change to increased information asymmetry in the market for REITs.
More recently, Buttimer, Hyland and Sanders (2005) analyze the long-term stock-market performance of REIT IPOs. In contrast to much of the literature on regular corporations, they find no evidence of abnormal post-IPO stock performance. Our study differs from these three in that our focus is on the degree to which the characteristics of the underlying "real" real estate markets (such as returns on unsecuritized commercial real estate, vacancy rates, the supply and demand for space and REIT dividend yields) can help explain REIT IPO volume, initial-day returns and long-term abnormal operating performance. Further, to the extent that we analyze post-IPO performance across cycles, we focus on operating performance rather than stock market performance. This has two advantages: It avoids the econometric difficulties in statistical tests of long-term, abnormal stock price performance, and it also evades the joint hypothesis problem inherent in attempting to disentangle abnormal stock market performance and market efficiency. In addition, to our knowledge, this study is the first to supply direct tests of the hypothesis that IPO waves are driven by investment demand. (2)
We expect our results to differ from those found for the average industrial IPO. Indeed, the REIT IPO market is not closely related to the broader IPO market. For example, in our data set, the contemporaneous correlation between REIT initial-day returns and average industrial IPO initial-day returns is -0.062. The correlation between the number of REIT IPOs and the number of industrial IPOs is 0.291.
One reason for this separation noted in point iv above is that the institutional structure of a REIT is very different than that of the typical corporation. These factors limit the degree of discretion a manager has by mandating a high dividend payout (3) and by restricting the types of activities in which a REIT may engage. These limitations reduce the manager's divertible free cash flow. An additional control on managerial discretion is the transparency of the underlying real asset market, which allows owners to more easily evaluate managers. Finally, as noted above, since REITs are a tax conduit, the tax shield from leverage is unavailable, taking away one of the most important benefits of debt versus equity issuance. This should decrease the cost of equity relative to debt, making equity issuance less of a negative signal of firm quality than it is for the typical corporation.
Collectively, these characteristics suggest that IPO models that are driven by informational asymmetries are less likely to be able to explain the behavior of REIT IPOs (compared to regular corporations). (4) But, due to this reduced role of information asymmetry and the transparency of the asset market, REITs should provide an especially powerful test of the link between underlying asset markets and IPOs. When interpreting our results, though, there is an open question as to the ability of one to generalize to the broader corporate IPO setting, especially in industries where information asymmetry is important.
To briefly summarize our key results, we find that both IPO volume (our measure of "hot supply" for REIT IPOs) and initial-day returns (our measure of "hot demand") are related to the state of the underlying real markets. We provide evidence that is consistent with the "heat" of the REIT IPO market being a function of both underlying markets and relative REIT prices. We also find that, while the performance of the real market is significant in explaining the supply-side definition of hot markets, dividend yield is both statistically and economically more significant and is the only variable significantly related to the demand-side definition of hot markets. Furthermore, future abnormal operating performance (as measured by return on assets and return on market equity relative to the median REIT) does not depend on whether the REIT IPO occurred in a hot or cold market. This conclusion is robust to our characterizations of hot or cold markets, implying homogeneous firm quality across IPO cycles. Finally, we demonstrate that lagged IPO proceeds are related to future sectorwide asset growth and to capacity utilization. These results collectively suggest that REIT IPO cycles are driven by the demand for investment capital rather than adverse selection.
The organization of the remainder of this study is as follows. The next section briefly overviews the relevant literature. The third section presents our data and our interpretation of it. The final section contains our conclusions.
IPO Literature
Theoretical Arguments
The theoretical and empirical IPO literature contains a variety of conflicting results. (5) The theoretical predictions most relevant to our research concern the quality of firms that issue equity during a "hot" IPO market. Hot IPO markets are generally characterized by a high volume of offerings, more initial-day underpricing and greater oversubscription. From one viewpoint, during hot markets issuers will attempt to exploit their informational advantage by timing their IPOs to take advantage of the market's overvaluation of their assets. An early example of this approach is Rock (1986). Rock's adverse-selection model features informed and uninformed investors competing for IPO shares. Uninformed investors do not know if issues are over- or underpriced and thus anticipate receiving shares when they are the least desirable. IPO underpricing thus arises as compensation for this winner's-curse phenomenon....
|
