Competition, Auctions & Negotiations in REIT Takeovers

The terms “mergers”, “acquisitions”, and “takeovers” are used interchangeably in this paper. Similarly, “target”, and “seller” are used interchangeably, as are “bidder” and “buyer”.

Although hostile takeovers occur infrequently in any industry, we simply point out that they occur even less in the real estate industry. Schwert (2000) finds that 25 % of his sample is characterized as hostile by the Securities Data Corporation (SDC). We perform a similar search in SDC, and find 4 % of REIT takeovers are classified as hostile. Furthermore, Schwert finds that most of the characteristics of hostile takeovers seem to actually reflect strategic bargaining behavior, and thus concludes that hostility may be quite a subjective classification.

The results and methodology from previous merger studies are summarized in Appendix Table 12.

In results not shown, the average time to completion for mergers in the sample (calculated as effective date minus announcement date, divided by 30) is 4.7 months, which is very similar to the 3.5 month average found by Allen and Sirmans (1987), Eichholtz and Kok (2008), as well as the 4.3 month average found by Womack (2012).

Real estate market cycles are known to be much longer than traditional macroeconomic business cycles, and tend to last in the neighborhood of 10 to 20 years (Geltner and Miller 2001).

Downs (1998) studied the adoption of this regulation within an event study framework and found that it produced a statistically significant and positive wealth effect for REIT shareholders.

The UPREIT structure essentially enabled properties to be acquired by UPREITs as a tax deferred like kind exchange, while the REIT Simplification Act of 1997 replaced the potential disqualification of REIT status with a penalty in situations where a REIT fails to follow certain IRS rules, modified the taxation of retained capital gains, repealed the 30 % gross income test, and made many other relatively more technical regulatory changes.

Howe and Jain (2004) find that REITs experienced positive wealth gains from the legislative events leading to the passage of The REIT Modernization Act of 1999. The Act provided two very important regulatory changes (in addition to several other relatively minor changes). First, it allowed a REIT to own a taxable REIT subsidiary (TRS) that can provide real estate related services without the REIT forfeiting its tax exempt status. Second, the REIT distribution requirement was changed from 95 % back to the 90 % level that applied to REITs from 1960 to 1980. The REIT Improvement Act of 2004 included a number of provisions to increase the operation efficiency of REITs. Most notably, changes included the allowance of REITs to make certain types of loans, effectively removing timber sales from the prohibited transactions tax, conforming the treatment of foreign shareholders to that of other publicly traded U.S. companies, and allowing REITs to avoid REIT status disqualification for non-intentional violations.

“Large” and “small” firms should be interpreted as relative sizes to a specific industry rather than as an absolute measure of total capitalization or assets. A large (small) firm in one industry might be quite small (large) in another industry.

We use the term “auction” throughout the paper to denote the auction-like process that is described in this section. While there are many similarities between this process and a classic/formal auction process, there are also notable differences. We use the term simply to convey that some target firms contact many potential buyers in anticipation of selling the firm to the highest bidder.

Our depiction of auctions versus negotiations as defined by the number of the number of potential bidders is supported by recent work from Paul Klemperer, who is highly regarded for his research on auctions. See for example Bulow and Klemperer (2009).

As a general rule of thumb, it appears that the average number of bidders involved in each phase of the merger process decreases by roughly half as the transaction reaches the next phase of the process.

“Non-REITS” denotes firms in all other industries, as most mainstream finance merger studies (including those cited above) exclude SIC Code 6798 (REITs).

Since we can observe only the means from the other studies, it is not possible to test for the statistical significance of the difference in means across the studies. However, given that the REIT averages are nearly always higher, even if the difference in means tests were insignificant, this would still imply that REIT merger competition on average is indifferent from other industries.

Brau et al. (2013) also document a substantial number of privitizations from the mid to late 2000s. See Ling and Petrova (2011) for a detailed study of the determinants of a publicly-traded REIT becoming a takeover target, and Brau et al. (2013) for a detailed study of the determinants of going private decisions.

Ling and Petrova (2011) find that all cash is used in 95 % of public-to-private REIT takeovers, although the study also finds that there has also been a shift toward the use of cash in public-to-public deals.

Which is the same estimation period used in the current study.

This study follows prior literature by placing primary reliance on the (−1,+1) window. This three day window is commonly utilized to compensate for imprecise measurement of when the merger announcement actually occurred. However, in results not reported, other event windows are utilized to ensure the reported results are robust to the choice of window specification.

Although a t-test of the difference of those means is statistically insignificant.

It should be noted that the greater transparency of REIT assets could somewhat reduce, but certainly not eliminate, the effect of the information cost hypothesis. The only case in which it would be eliminated would be in a world in which all information for REITs is already reflected in share prices. Obviously, this world currently does not exist. Rather, we are likely in a world of reduced information asymmetries, where more information (particularly in regards to value) is known about real estate assets than assets in other industries. Furthermore, the dramatic increase in the use of an auction sales process by REIT targets in Period 2, which is highly correlated with the growth of private equity, indicates a change in the sales process chosen by REITs based on factors external to the firm.

Specifically, ln(Assets), Market2book, and ln(Years Old), which collectively proxy for firm size and market exposure, have explanatory power in the choice of sales procedure (see the first-stage of Model 1) but do not have explanatory power for merger returns (see the first-stage of Model 2). Therefore, these variables are omitted in the second-stage of Model 1. Similarly, Equity REIT and Hybrid REIT have explanatory power for merger returns (see the first-stage of Model 2) because different asset classes have different returns, but the variables do not have explanatory power for the choice of sales process (see the first-stage of Model 1). Therefore, these variables are omitted in the second-stage of Model 2. It should be noted that that Public, Affiliated, Cash, Unsolicited, and Early Announce are omitted from the model of sales procedure since these variables occur after the sales procedure decision is made.

In contrast, the initial public offering literature (IPO) has examined this issue more in-depth, as REIT IPOs also exhibit lower returns than IPOs in other industries (that is, REIT IPOs have lower underpricing, and therefore lower returns). For example, Gokkya et al. (2013) and Steele et al. (2013) present arguments that there is less unknown public information, or information asymmetries, regarding real estate assets. These studies suggest that asymmetric information in REITs should be relatively minor because real estate investors use similar conventions in the property valuation process, REITs face frequent property value assessment by governmental agencies, and REITs pay regulated, stable dividend payments. Therefore, if other firms are able to value REITs with comparatively greater ease, then the associated asymmetric information cost should be relatively less than when acquiring companies in other industries.