|
Article Excerpt
This study investigates the relation between firms' investment choices and various governance mechanisms, using a sample of real estate investment trusts (REITs). We find evidence that the responsiveness of REITs' investment expenditures to their opportunities depends on their corporate governance structures. Within the set of governance mechanisms that we examine, we find particularly strong links between investment behavior and ownership. Specifically, we find that the investment choices of REITs are more closely tied to Tobin's q if they have greater institutional ownership or if they have lower director and officer stock ownership. These results are consistent with institutional owners monitoring the firm's investment policies as well as with high insider ownership allowing managers to follow their own investment agendas.
**********
Manager-shareholder agency problems arise when managers' interests diverge from those of shareholders, resulting in lower firm values (Jensen and Meckling 1976). Several governance mechanisms are devised to alleviate these problems, such as the monitoring role of the board of directors and large shareholders, and incentive alignment achieved through managerial ownership and compensation. But the effectiveness of these mechanisms remains an open question.
To address this issue, researchers have attempted to measure the relation between firm performance and various governance variables. For example, Yermack (1996) documents that firms that have smaller boards--which he argues are more effective in monitoring managers--tend to have higher values. Another well-known example is Morck, Shleifer and Vishny (1988), who find that firm values initially increase with increased managerial ownership, as a result of convergence of interests between managers and shareholders, then decline as the managers become more entrenched and finally increase slightly as managerial ownership increase further. In addition to these cross-industry studies, there are a number of studies that examine the effect of governance on the performance of REITs. For example, Capozza and Seguin (2000, 2003) find that REIT valuations are increasing in both insider ownership and the use of self-administered management structures.
These cross-sectional studies, however, are often fraught with endogeneity problems, as pointed out by Himmelberg, Hubbard and Palia (1999), Coles, Lemmon and Meschke (2003) and others. For example, more successful managers are likely to have more influence on their firms' governance structure, inducing a positive correlation between firm values and management-friendly governance structures. Alternatively, Tobin's q, a common measure used in these studies (see Tobin 1969), may be partly capturing elements that differ across firms but are not related to performance. For example, firms with intangible assets, which may have special governance needs, tend to have high Tobin's qs. The latter problem provides some motivation for using REITs as a laboratory, because focusing on a single, homogeneous industry mitigates these unobserved differences.
This article contributes to the governance literature in two ways. First, we follow Capozza and Seguin (2000, 2003), among others, and focus on real estate investment trusts (REITs), which provide more reliable measures of Tobin's q, and thus provide a cleaner experiment for examining the effect of governance on firm performance. Second, in addition to considering the effect of corporate governance on firm performance, we consider how governance influences corporate decisions--in particular, the level of investment expenditures.
Investment is a natural choice to consider owing to the tension it creates between managerial and shareholder interests. If left to their own devices, managers have a tendency to "empire build" (e.g., Baumol 1959, Donaldson 1984, Jensen 1986), which can be thought of as investing in negative net present value (NPV) investment projects to pursue growth as an objective rather than value maximization. (1) Running a larger firm can lead to greater opportunities to extract private benefits, more prestige for the CEO and greater compensation (e.g., Murphy 1999). As a result, CEOs have personal incentives to invest on behalf of the firm in a way that is not necessarily sensitive to the opportunities that the firm faces. In contrast, shareholders would prefer that the CEO only invest when it is positive NPV, that is, when it is warranted given the firm's opportunity set. Corporate governance mechanisms, then, can help reconcile this conflict.
Our empirical tests are based on the idea that well-governed REITs tend to invest more when their investment opportunities are better, while the investment choices of REITs that are managed in the interests of empire-building managers rather than shareholders will be less sensitive to changes in investment opportunities. Specifically, we conjecture that the investment expenditures of REITs with stronger corporate governance structures will be more closely tied to the average Tobin's q of REITs in their respective property sectors.
To measure the strength of corporate governance, we focus on several of the most important mechanisms that have attracted considerable attention: stock held by institutions, blockholders and insiders, managerial compensation, the size and the composition of the board of directors as well as the state of incorporation. In addition to their use in the academic literature, many of these variables are used as indicators of a firm's governance structure by real estate industry professionals (see Starkman 2005). However, these mechanisms are especially vulnerable to the endogeniety issue mentioned earlier. (2) For example, firms that are performing well would likely pay their managers more and attract greater interest from institutional investors.
Testing for a relation between investment and governance is difficult in a broad corporate setting due to the difficulty in measuring investment opportunities across firms and the potential sensitivity of investment to cash flow. In this regard, our sample of REITs provides us with several advantages, including greater transparency (and, therefore, better measures of Tobin's q) and payout restrictions that lessen the ability of cash flow to impact investment. Because of their unique characteristics, REITs should be able to provide more powerful tests of our hypothesis than most other corporations.
We find some evidence of time series relations between governance and value within firms. But, in contrast to earlier studies on broader samples of firms, we find little evidence of a reliable cross-sectional relation between our measures of corporate governance and firm performance, as measured by Tobin's q. One interpretation of this finding is that there is very little cross-sectional dispersion in governance structures within any particular industry like REITs, which is why a cross-industry study is required to detect a relation between governance and performance. A second interpretation is that cross-industry differences in Tobin's q capture intangible sources of value that may be spuriously correlated with governance. Finally, it is possible that the reverse causality, running from performance to governance, is especially acute for REITs, which would bias the tests against finding anything. This reverse causality could also generate the observed pattern in within-firm variation.
Our examination of the effect of governance on investment expenditures is not likely to be subject to the same endogeneity concerns as our tests that examine the effect of governance on firm values. But these tests also require sufficient cross-sectional dispersion in governance. However, our results do in fact find a significant relation between governance and REIT investment choices, which suggests that the lack of power that arises because of insufficient cross-sectional dispersion in the corporate governance is not too severe.
In particular, we find that the investment choices of REITs with greater institutional ownership tend to be more sensitive to changes in investment opportunities. This is consistent with the idea that institutional investors monitor REITs' investment decisions, leading them to increase (decrease) their growth when they face better (worse) opportunities. In addition, we find that the investment choices of REITs are more sensitive to investment opportunities when directors and officers own fewer shares, which is consistent with the hypothesis that more entrenched managers tend to follow their own agendas rather than investing when the opportunities are the most attractive.
We also separately investigate changes in equity and debt to determine if the effects on investment are attributable to one financing channel. We find that the observed effects of governance are not driven by one source of capital. Director and officer ownership has similar relations with the sensitivities of changes in both equity and debt to q. Board independence, in contrast, is more closely linked to the sensitivity of equity changes, while institutional ownership is significantly related to the sensitivity of changes in debt to investment. This is consistent with entrenched managers investing more heavily in the face of poor investment opportunities via debt and equity issuance, and with firms with high institutional ownership finding it difficult to raise additional debt in bad times.
Our article is organized as follows. The next section provides background information on REITs, and it discusses the implications of this organizational form for our study. We then further develop our hypothesis and test design. Next, we discuss our data, followed by a summary of our empirical results. The final section concludes.
REIT Features and Implications
Using REITs to test for the impact of governance on firm decisions presents some distinct advantages. First, as we briefly mentioned, because of their tangible assets and relatively transparent structures, REITs provide better measures of Tobin's q. Unlike many firms whose most significant assets are off their books (e.g., human capital or a technological advantage), REITs' values are derived primarily from physical assets that are reflected in their financial statements. While we estimate q using book values, because of the breadth of our sample, Gentry and Mayer (2003a, 2003b) note that, for the largest REITs, Tobin's q can be more accurately estimated with appraisal data. (3)
Second, while previous empirical studies document that internal cash flow is an important determinant of investment choices, (4) REITs are required to pay out almost all of their earnings as dividends to be exempt from corporate taxes. (5) So, while we control for cash flow in our tests, we expect less heterogeneity in REITs' ability to respond to the investment opportunities that arise because of differences in internally generated cash, which should help the power of our tests.
This dividend-payout requirement also implies that institutional investors may play an especially important role in monitoring REITs. Because of the forced payout, REITs must frequently return to the capital markets for new external funds. This process of raising capital provides outsiders with an opportunity to collect information and scrutinize the firm, its prospects and its performance. (6) Gibson, Safieddine and Sonti (2004) analyze the behavior of institutional investors around firms' seasoned equity offerings (SEOs) and find evidence that institutions use the information collected around SEOs to help discern firm quality (i.e., institutions appear to be able, ex ante, to discriminate between firms that exhibit good or bad post-SEO performance). Thus, REITs' frequent capital market activities may contribute to the extent to which they are monitored by institutional investors.
Another advantage of using REITs arises from a second requirement for firms to maintain REIT status: at least 75% of the assets of a REIT must be real estate related, and at least 75% of its gross income must be derived from real estate rents or interest on mortgages on real properties. This requirement reduces heterogeneity in investment opportunities within the same property-type market. Because REITs are vehicles for real estate ownership, two REITs that invest in the same property type should face very similar opportunities. Accordingly, in our empirical tests, we use q for each property type for each year as our measure of investment opportunities, rather than a firm-specific measure. By averaging across firms within each property type and year, and by measuring the shared component of q, we reduce measurement noise due to firm-specific idiosyncrasies. (7)
A third important requirement of REIT status is related to breadth of ownership. (8) No more than 50% of a REIT's shares can be owned by five or fewer shareholders, and a REIT must have at least 100 shareholders (the "five or fewer" rule). After 1993, institutional investors are not counted as a single owner; instead, their ownership is passed through to their beneficiaries for the purposes of meeting this requirement. This change has led to an increase in institutional ownership of REITs (e.g., see Downs 1998), which could provide another source of monitoring for REIT management. (9)
In addition to these advantages, REITs have substantial variation in both governance mechanisms and investment opportunities. Many governance mechanisms were introduced into REITs in the 1990s. Before 1990, most REITs contracted the management function to outside advisors. (10) Now, the vast majority of REITs are like general C corporations; REITs have boards of directors and their managers own REIT shares. In fact, data from recent surveys conducted by Institutional Shareholder Services of governance practices indicates that the incidence of most governance mechanisms is similar for real estate firms compared to other publicly traded corporations. (11) Furthermore, since the early 1990s, there has been an increased presence of large shareholders and institutional investors.
One alternative disciplining device that is weaker in the REIT sector is the threat of hostile takeovers. (12) Chan, Erickson and Wang (2003) suggest that this is because many REITs use the so-called excess shareholder provision, making it difficult for acquirers to acquire a large equity position in REITs. The virtual absence of takeovers suggests that the monitoring role of the board and large shareholders and the incentive effects of managerial ownership and compensation are likely to have a more important effect on the value of REITs.
In addition to the heterogeneity in governance choices, the cyclical nature of this industry implies that real estate investment opportunities...
|
