How Does Corporate Governance Affect the Implied Cost of Equity Capital? Evidence from REITs

Transcription

1 How Does Corporate Governance Affect the Implied Cost of Equity Capital? Evidence from REITs Tom Thibodeau Leeds School of Business Ying Xiao* Mount Saint Mary College University of Colorado, Boulder, CO Powell Ave, Newburgh, NY Abstracts: REIT provides us an excellent laboratory to study whether legal and regulatory constraints serve as effective substitutes for good corporate governance. This study examines the relationship between the cost of equity capital and internal governance in REITs. We find that REITs have a much lower cost of equity compared to unregulated industries. Measures of Internal REIT corporate governance such as value of directors shareholdings, and the percentage of directors with tenure are negatively associated with the excess cost of equity. The relationship is only significant in small REITs and small REITs with more free cash flow, consistent with the notion that corporate governance matters more when the agency problem is more severe. *Corresponding author. Address: School of Business, Mount Saint Mary College, 330 Powell Ave, Newburgh, NY, 12550, tel: , ying.xiao@msmc.edu. 1

2 I. Introduction Although current studies have concentrated on corporate governance on nonregulated industries; the issue has gained considerable currency for REITs as well. The regulatory environments of REIT have made it an excellent laboratory to study whether legal and regulatory constraints serve as effective substitutes for good corporate governance. Even though it seems that regulation helps to protect investors rights, whether good corporate governance still matters is an empirical question. REITs are different from the other industries mainly because they are subject to four categories of requirements regarding restrictions on assets, income, distribution, and ownership structure. First, they are required to distribute at least 90% of taxable income to shareholders. With limited internal capital, REIT managers have to rely heavily on outside financing through either debt or equity. Unlike other taxable firms, debt is a less attractive alternative. Second, at least 75% of REITs income should be derived from real estate investments. As REITs mainly involve in real property transactions that include a wide range of heterogeneous, illiquid assets, it is difficult for shareholders to determine the fair market value of these transactions. This results in monitoring REITs managers critical. Third, REITs cannot hold more than 25% of assets that are not associated with real estate. This limits REITs managers exposure to industries other than REITs and limits their job market potentials. For their career concerns, they are more likely to collude to avoid hostile takeover threats. Finally, REITs must maintain a diversified ownership with at least 100 shareholders, the five biggest of which may not own more than 20 percent of the total shares outstanding. The lack of concentrated 2

3 ownership diminishes the effectiveness of monitoring by the market for corporate control, and exacerbates the lack of transparency. As Eicholtz and Kok (2008) point out, in 95 property takeovers they studied, only 2 were hostile in nature. Campbell et al. (2001) and Bianco et al. (2007) also document the absence of hostile takeovers in REITs. As mentioned above, the absence of hostile takeover in REITs has made many researchers believe that external governance mechanism is ineffective for REITs. Thus, the efficacy of internal governance is critical for shareholders. Bianco, Ghosh and Sirmans (2007) find low G-index and the irrelevance of G-index in more recent times. Campbell, Ghosh, Petrova and Sirmans (2009) find evidence supporting that antitakeover defense measures have reduced importance for REITs. Although researchers have studied the relationship between corporate governance and the valuation of REITs, the channel through which corporate governance influence REIT firm value is missing. Ghosh and Sirmans (2003) find weak relationship of board independence enhances REIT performance. Feng, Ghosh and Sirmans (2005) find that the effect of good governance on performance is significant only for the best and worst boards. Our study thus fills in the gap by examining the denominator effect-cost of equity and provides one channel to connect the relationship between corporate governance and REITs valuation. To the best of our knowledge, we are the first to examine the relationship between corporate governance and cost of equity for REITs. To compute cost of equity, we employ ex-ante analyst forecast data. Recent finance and accounting literature has documented the problem of using ex-post returns as proxies of cost of equity. Hail and 3

4 Leuz (2006a, 2006b), Stulz (1999) point out that ex-post returns not only capture differences in a firm s cost of equity, but may also reflect the shocks of a firm s growth opportunities, differences in expected growth rates, and changes in investors risk aversion. They argue that using ex post returns to estimate expected returns thus requires average realized returns over a long period. Elton(1999) argues that even over a long period, however, the realized returns may differ from expected returns. Furthermore, Fama and French (1997) also suggest that ex-post returns are poor proxies for the cost of equity. The ex-ante cost of equity, however, makes an explicit control for cash flows and growth potential (Hail and Leuz, 2006a) and may provide a better measure for the expected return (Pastor, Sinha & Swaminathan (2008). After 2001, analyst forecasts coverage for REITs increases, which makes it possible to compute the ex-ante cost of equity for REITs. The number of analyst forecasts for REITs ranges from 25 to 74 prior to the year 2000 and increases to 120 in year 2004 (see Devos, Ong and Spieler (2007) table 1). In 2001 REIT industry experienced some key structural changes and has thereafter attracted more analyst coverage is a landmark year when six REITs were added to one of the S&P general market indices and the required dividend distribution for REITs fell from 95% to 90% that year. Devos, Ong and Spieler (2007) show that analyst coverage gets more accurate and less biased after year We employ ex-ante cost of equity in this paper and find that weak internal corporate governance is associated with higher cost of equity for REITs. In general, board size, percent of tenured directors, the value of directors share holdings, and audit committee independence are significantly negatively associated to excess cost of equity 4

5 in REITs. The association between internal corporate governance and cost of equity is significant only for small REITs and small REITs with more free cash flow, consistent with the hypothesis that corporate governance matter more when there is more severe agency problems. Our study contributes to the corporate REITs governance literature by indentifying a channel through which governance can improve firm value. Ghosh and Sirmans (2005) find that REITs with weaker board monitoring have higher CEO compensation. Campbell, Ghosh, Petrova dna Sirmans (2009) report that REITs acquirers with weaker internal governance experience lower announcement-period abnormal returns. While all of these studies show that weaker governance is associated with lower expected cash flow (the cash flow effect), our study suggests that weaker governance is also associated with higher discount rates (the discount rate effect). More important, we also indentify that governance matters more in reducing the cost of equity when agency problems from free cash flow are more severe. Our study also extend the literature on REITs information asymmetry and corporate governance by providing evidence that good corporate governance reduce the cost of raising funds. Anglin et al (2010) show that good corporate governance reduce information asymmetry by reducing the bid-ask spread in REITs. We present a direct benefit of good governance for REITs. II. Hypothesis Development and Model Specification. Unlike non-regulated industries, there is compelling evidence that the takeover market is ineffective in the REIT sector. Campbell et al. (2005) and Eicholtz and Kok 5

6 (2008) establish that hostile takeovers are virtually non-existent among REITs. The low G-index and the irrelevance between G-index and REIT performance (Bianco, Ghosh and Simans 2007) suggest that given the ineffectiveness of external governance in REITs, more attention should be paid to internal governance in REITs. In this paper, we focus on the effects of internal governance upon cost of equity. Internal governance mechanism can reduce the cost of equity in several ways. First, corporate governance can reduce the non-diversifiable risk of expropriation by corporate insiders (la Porta, Lopez-de-Silanes, Shleifer and Vishny, 2000). When the macroeconomics condition is experiencing a downturn, insiders are more likely to expropriate more. At the same time economic downturn dries up firms investment opportunities. This negative relationship between insider expropriation and market conditions magnify the systematic risk of a firm which must be compensated by a higher required rate of return. Second, corporate governance lowers the cost of equity by reducing the cost of external monitoring by outside investors (Lombardo and Pagano 2002). If outsiders have to spend more time and resources on monitoring firms management because of poor corporate governance, they will accordingly demand a higher required rate of return. Third, corporate governance also reduces the cost of equity by limiting optimistic trading and thus diminishing information asymmetry (Easley and O Hara 2004). Anglin et al (2010) find that better governance reduce information asymmetry as measured by the percentage bid-ask spreads demanded by the market in REITs. 6

7 H1: Since external governance is weak in REITs, internal governance mechanisms are critical in monitoring. There is a negative relationship between internal governance and the cost of equity in REITs. H2: When agency problem is more severe, the relationship between internal governance and the cost of equity is more significant. We investigate the effect of internal corporate governance on the excess cost of equity by regressing the excess cost of equity on a comprehensive set of corporate governance measures, controlling for other non-governance variables. R i,t = α + βcg i.t + γcontrols i.t + ε i,t Firm level corporate governance (CG) is measured by a wide range of the structure and activities of the board of directors and its audit committee. All the governance variables are from the Corporate Library database. Similar to prior research in corporate governance, we examine nine governance variables that include three categories: board structures, director attributes and audit committee attributes. For board structure, we examine (1) board_ind: board independence, defined as the percentage of fully independent directors on the board; (2)board_size: board size, measured as the number of directors, (3)board_meet: board meetings, measured as the number of full board meetings held per year. For directorships, we examine (1)dir_tenure: the percentage of directors with tenure exceeding 15 years; (2)dir_exp: the percentage of directors with more than 4 concurrent corporate (public) directorships; (3)dir_comp: the reported directors base compensation; (4)dir_share: values of company stock owned by 7

8 the board of directors. For audit committee attributes, we examine (1) audit_ind: audit committee independence, equal to 1 if the audit committee is comprised wholly of independent directors and 0 otherwise; (2) audit_exp: the number of audit committee members who are designated financial experts. We control for several risk factors and firm characteristic that affect the cost of equity. Specifically, we control for cross-firm differences in beta, firm size, book-tomarket ratio, price momentum, analysts earnings forecast errors, liquidity, and the fixed effects of year. The capital asset pricing model (CAPM) predicts that the cost of equity is positively associated with the market beta. We calculate beta based on the previous five years monthly stock return (a minimum of eighteen months data are required to calculate beta). Beta is calculated as in EXRET 0 1RMRF where EXERT is the firm's monthly return minus the risk free rate, RMRF is the excess return on the market. Fama and French (1992) find that firm size and book-to-market ratio explain the cross section of stock returns. Firm size is measured as the natural log of fiscal year end market value of equity (COMPUSTAT csho*compustat prcc_c). The market to book ratio is calculated as fiscal year end market value of equity divided by fiscal year end book value of equity. Following Guay et al. (2003), we include price momentum to mitigate biases in the estimation of cost of equity driven by analysts sluggishness with respect to information in past stock returns. Price momentum is measured as the compounded return over the previous six months. Following Hail and Leuz (2006), we include analysts 8

9 earnings forecast errors to control for the potential effect of analysts biases on the estimation of the cost of equity. Analysts earnings forecast error is measured as actual earnings minus the forecast consensus for the next year scaled by the current year s stock price. As Amihud and Mendulson (1986) points out that in equilibrium, expected returns are higher for illiquid stocks. To control for liquidity we include the bid-ask spread. It is measured as the average of the daily bid-ask spread in the previous year before the I/B/E/S forecast reporting date. Daily bid-ask spread is measured as the difference between the closing bid and ask prices scaled by the average of the closing bid and ask prices and multiplied by 100. For firms that do not have 252 trading day data available, we use all the available days that has trading data. Only 7 REITs in our sample do not have 252 trading days, from those the number of trading days ranges from 112 to 225 days. III. Data and Descriptive Statistics. 1. Sample Selection. We derive our data from several databases. Our initial sample consists of all REITs with corporate governance data from The Corporate Library database. We first match these firms to Compustat to retrieve the book value of equity, net income, dividends, total assets at the fiscal year end before proxy statement date. We then match the sample with I/B/E/S to retrieve analysts earnings forecasts to estimate the implied cost of equity. For REITs that have proxy dates between January and March, we choose month +4 forecasts. For REITs that have proxy dates on or after April, we choose the first 9

10 forecast month after the proxy statement date. This methodology makes sure that the expected earnings (and computed cost of equity) encompass the released information of corporate governance. Finally, we eliminate the observations with missing control variables. Our final sample includes 251 REITs in the United States between 2003 and Table 1 shows the sample selection process and distribution of the firm-year observations. [insert Table 1 here] 1. Measure of firm level corporate governance. Table 3, Panel C contains descriptive statistics on corporate governance variables. The sample REITs have an average of 8.69 board members, and hold an average of 7.94 board meetings per year. On average, 66 percent of board members are independent. Seven percent of directors hold directorships on more than 4 corporate boards and seven percent have been on the board for at least 15 years. On average, value of shares owned by a director is $34 million percent of REITs have audit committees fully composed of independent directors. On average, 13 percent of members per audit committee are designated financial experts. [insert Table 2 here] 2. Estimation of the cost of equity. The independent variable: cost of equity is computed using four models (Gebhardt et al. (2001), Claus and Thomas (2001), Easton(2004), Ohlson and Juettner- Nauroth (2005)). Since there is no consensus on which model performs the best, we 10

11 follow (Chen et al and Chen et al. 2010) to use the median of the estimated cost of equity computed by these four models as our estimate of the cost of equity in REITs. See appendix A for the detailed description of the four models estimating the cost of equity. [Note: current result is based on one method of calculation, will need to calculate more or justify which method to choose.] The estimated excess cost of equity for REITs is 1.8%, much less than other industries in general. For example, Chen et al. estimate the excess cost of equity for non regulated industries and they find the excess cost of equity is 4.9%. 3. Control variables. Panel D of Table 2 provides the summary statistics for our control variables. The mean of BETA is 0.45, which suggests that the average systematic risk of our sample REITs is lower than that of the firms in the value weighted market index. The mean FERR is , indicating the optimism of analysts. The mean liquidity measured as the bid-ask spread is Compared with Anglin (2010), our sample firms show better liquidity, which is expected given that fact that the final sample REITs are, on average, larger than their sample REITs. The mean market value of equity is US$3.4 billion, which indicates that our sample REIT is relatively large. This is not surprising, since I/B/E/S tends to cover large firms. 2. Main Results To compare with non-regulated industry, we first examine earnings forecast errors for REITs based on the market value of REITs and the number of analysts. We found 11

12 that the average forecast error is -0.6%, lower than -1.04% of unregulated industries found in Chen et al (2010). This is not surprising given that REITs have more predictable earnings since the rent increase is more reliable once the macro economy data are available. When separating REITs into small, medium and large REITs based on the standard of small-cap, mid-cap and large-cap firms, we find that the size of REITs is negatively corrected with analyst forecast errors. The absolute value of analysts forecast error is 2.06% for small REITs and 0.175% for big REITs. The number of analysts does not seem to reduce forecast errors in general. Analysts forecasts are more optimistic for small REITs compared to large REITs. The forecast error for small REITs is -12%; for large REITs this number is 0.3%. Compared to non-regulated industries, REITs have a lower cost of equity on avearge. The average REITs in our sample has an excess cost of equity of 1.8% compared to 4.9% in regular industries; this is consistent with the notion that regulated industry has less agency problem, especially because REITs have a much higher payout rate compared to regular firms. Now we turn to the relationship between corporate governance and the cost of equity. Table 3 shows the OLS regression results of excess cost of equity on internal governance variables. For the entire sample (column (1)), board size is negatively correlated with the excess cost of equity. One extra member in the corporate board decreases the cost of equity by -0.3%, which is 1/6 decrease from the mean. REITs with 12

13 payout ratio greater than one have 1.1% lower cost of equity. Most of other independent variables show the negative sign but not significant. [Insert Table 3 here] Table 3 column (2) shows the results for small REITs. The percentage of directors with tenure exceeding 15 years is significantly negative. This shows that a 10 percentage increase in the percentage of the number of tenured directors decreases the cost of REIT equity by 0.58%. This shows that the effectiveness of a board may depend upon its experience. The value of directors shareholdings is negatively correlated with the cost of equity. An extra one million increase in the value of directors shareholdings is associated with 0.1% lower cost of equity. This is consistent with directors compensation is positively related to monitoring activity by directors. The independence of audit committee independence can effectively reduce the cost of equity by 5.6% as shown in column (2). The results in column (3) of Table 3 are for small REITs with payout ratio less than 1. The results are similar to column (2) that board size, percentage of directors with tenure exceeding, the value of directors shareholdings, and the audit committee independence are negatively correlated with excess cost of equity in REITs. The results in Table 4 add control variables besides those independent variables in table 3. Similar to Table 3, corporate governance variables are more important for small REITs and especially small REITs with payout ratio less than 1. For large REITs, internal corporate governance does not seem to reduce cost of equity. [Insert Table 4 here] 13

14 3. Future direction. To compare the REITs industry with other industries, we would like to further examine the G-index in REITs with the cost of equity. While Chen et al. (2010) document that G-index is significantly related to cost of equity in non-regulated industries, REITs provides a tests of whether the effect disappear when it is claimed that external governance is not effective. 14

15 References Anglin, Edelstein, Gao and Tsang, "How does Corporate Governance affect the Quality of Investor Information? The Curious Case of REITs," Journal of Real Estate Research, forthcoming. Bianco, Ghosh, and Sirmans, The Impact of Corporate Governance on the Performance of REITs. Sep 2007, Journal of Portfolio Management. Campbell, Ghosh, Petrova and Sirmans, Corporate Governance and Performance in the Market for Corporate Control: The Case of REITs, 2009, The Journal of Real Estate Finance and Economics. Chen, Chen, and Wei, Legal protection of investors, corporate governance, and the cost of equity capital 2009, Journal of Corporate Finance, Chhaochharia and Grinstein, CEO Compensation and Board Structure, 2009, Journal of Finance. Claus and Thomas, Equity premia as low as three percent? Evidence from analysts' earnings forecasts for domestic and international stock markets, 2001, Journal of Finance 56, Devos, Ong, and Spieler Analyst Activity and Firm Value: Evidence from the REIT Sector, 2007, Journal of Real Estate Finance and Economics. Easton, PE ratios, PEG ratios, and estimating the implied expected rate of return on equity capital, 2004, Accounting Review 79, Gebhardt, Lee, and Swaminathan, Toward an implied cost of capital, 2001, Journal of Accounting Research 39, Ghosh and Sirmans, Board Independence, Ownership Structure and Performance: Evidence from Real Estate Investment Trusts, 2003, Journal of Real Estate Finance and Economics. Ghosh and Sirmans On REIT CEO Compensation: Does Board Structure Matter?, 2005, The Journal of Real Estate Finance and Economics. Hartzell, Sun and Titman, The effect of Corporate Governance on Investment: Evidence from Real Estate Investment Trusts (REITs), 2006, Real Estate Economics. Ohlson and Juettner-Nauroth, Expected EPS and EPS growth as determinants of value, 2005, Review of Accounting Studies 10,

16 Appendix: 1. Cost of equity estimated from the Model of Gebhardt et al. (2001) T 1 P t = B t + FEPS t+i r e B t+i 1 (1 + r e ) i + FEPS t+t r e B t+t 1 r e (1 + r e ) T 1 i=1 2. Cost of equity estimated from the Model of Claus and Thomas (2001). 5 P t = B t + FEPS t+i r e B t+i 1 (1 + r e ) i + (FEPS t+5 r e B t+4 )(1 + g lt ) (r e g lt )(1 + r e ) 5 i=1 3. Cost of equity estimated from the Model of Easton(2004). P t = FEPS t+2 FEPS t+1 + r e FEPS t+1 POUT r 2 e 4. Cost of equity estimated from the Model of Ohlson and Juettner-Nauroth (2005) (According to the Procedures in Gode and Mohanram(2003)) P t = FEPS t+1 + FEPS t+2 FEPS t+1 + r e FEPS t+1 (1 POUT) r e r e (r e g lt ) Which suggests that Where r e = A + A 2 + FEPS t+1 ( FEPS t+2 FEPS t+1 g P t FEPS lt ) t+1 A = 1 2 (g lt + POUT FEPS t+1 P t ) This appendix details the variable construction for analysis of the Compustat sample. All numbers in parentheses refer to the annual Compustat item number. Total Debt=short-term debt (DLC)+long-term debt (DLTT). Book Leverage=total debt/book assets (AT). Market Equity=stock price(prcc_f)*shares outstanding (CSHPRI). Market Leverage=total debt/(total debt+market equity). 16

17 17

18 18

19 19

20 20