Corporate Governance, Cost of Capital and Performance: Evidence [PDF]

Corporate Governance, Cost of Capital and Performance: Evidence from Australian Firms Peter Kien Pham School of Banking and Finance University of New South Wales Jo-Ann Suchard* School of Banking and Finance University of New South Wales Jason Zein School of Banking and Finance University of New South Wales

Using a sample of large Australian firms from 1994 to 2003, we show that variation in firm-level corporate governance mechanisms plays an important role in explaining a firm’s cost of capital. Our empirical results show that greater insider ownership, the presence of institutional blockholders and smaller and independent boards all serve to reduce the perceived risk of a firm, thereby leading investors to demand lower rates of return on capital provided. This highlights the important role that corporate governance plays in creating value for shareholders by reducing the cost of external financing. Given the inconclusiveness of existing literature that uses Q to measure firm value, this research provides an alternative and potentially more suitable way to investigate the impact of corporate governance on firm value.

JEL classification: G32 Keywords: Corporate governance; Firm value; Cost of capital; Australia

_____________________________ *Corresponding author, School of Banking and Finance, University of New South Wales, Sydney NSW 2052 Australia, Tel. +612 9385-5876, e-mail: [email protected]

Electronic copy available at: http://ssrn.com/abstract=1015986

1. Introduction The role of corporate governance in creating value for shareholders has become the subject of intense interest in corporate finance research. From the early work of Jensen and Meckling (1976), Demsetz and Lehn (1985), Shleifer and Vishny (1986), the theoretical and practical importance of mechanisms that align the interests of managers and shareholders as well as those that curb ‘insider’ expropriation have been widely acknowledged. However, despite this general acceptance of the role of corporate governance, empirical research has remained inconclusive regarding the extent to which individual monitoring mechanisms enhance firm performance and shareholder value. In particular, previous attempts to investigate the relation between the strength of corporate governance and firm value have not convincingly overcome two critical difficulties: the potential endogeneity associated with monitoring mechanisms and the lack of an accurate and stable measure of performance.

In this study, we examine the value-creation role of corporate governance mechanisms using an alternative approach to those used in most previous studies. Firstly, rather than measuring firm value directly using variables such as Tobin’s Q, we investigate the relation between a firm’s governance mechanisms and its cost of capital. While most previous studies focus on the fact that a strong governance environment can limit divergence of cash flows, we argue that it can also reduce the cost of capital (and hence, increase firm value indirectly). A firm’s cost of capital reflects investors’ required return based on the firm’s systematic risk. A number of possible risks arise when corporate governance is weak. As external monitoring becomes more difficult, insiders may not pursue value maximizing strategies, instead opting for strategies that entrench their positions. For example, excessive borrowings

1 Electronic copy available at: http://ssrn.com/abstract=1015986

and empire building expansions are typical self-serving activities that also increase a firm’s exposure to market-wide risk and ultimately, increase the cost of capital. Furthermore, weak governance often results in a lack of corporate transparency, which translates into higher issuing and transaction costs. This increases a firm’s cost of capital even further.

Secondly, to control for the endogeneity of our governance measures, we employ a fixed-effects regression model and use a sample of large Australian firms from 1994 to 2003. Himmelberg et al. (1999) argue that corporate governance mechanisms are not entirely predetermined, but may reflect the agency-cost and contracting environments of a firm. As a result, corporate governance and firm value may be driven by common firm characteristics, some of which are neither clearly observable nor measurable. For example, managers tend to hold large ownership stakes (which is commonly viewed in the literature as a mechanism to combat agency problems) in high-risk and high-growth firms to signify their commitment. Further, with the use of equity-based remuneration, insider ownership may automatically increase after periods of strong performance. However, this spurious correlation does not offer any insight into the impact of insider ownership in reducing agency problems and improving firm value.

The Australian corporate system offers a relatively unique environment to assess the impact of corporate governance mechanisms on the cost of capital. Australian firms have board structures and mechanisms that are similar in design to Anglo-Saxon boards and are in contrast to German/Japanese boards. In addition, the Australian market is not a bank-centred market, in which banks take an active role as an equity holder and corporate monitor, as in Germany and Japan. However, compared to the

2

US and the UK, the Australian market for corporate control is much less active as a corrective mechanism against management entrenchment and corporate failure. This makes the role of internal governance mechanisms such as independent boards, and management incentives more important in Australia.

We find that variations in firm-level governance characteristics, such as board independence and size, the presence of institutional blockholders and insider ownership significantly affect a firm’s cost of capital and thus implicitly enhance firm value. This approach differs to and has several advantages over those of previous studies, which often employ the Tobin’s Q measure of firm value. First, the cost of capital is a much more stable measure than previously employed proxies for firm value, and hence, our inferences are less subject to errors. We replicate tests from past research that have attempted to establish a link between firm performance (as measured through Tobin’s Q) and governance mechanisms (mainly managerial ownership). However, we do not find a significant relationship, thereby highlighting the significance of our alternative approach.

Second, the fixed-effects regression model provides a useful tool to address the potential endogeneity problem associated with various corporate governance characteristics. Other studies apply the fixed-effects regression methodology but their results are often insignificant, and hence difficult to interpret due to the large intertemporal variations of firm value/performance measures. Our focus on cost of capital highlights that strong corporate governance can reduce a firm’s systematic risk and information asymmetry, in addition to the role of limiting cash flows divergence, as suggested in past research. Overall, even though firm value cannot be measured

3

directly and accurately, our results suggest that corporate governance can influence a firm’s value indirectly through its cost of capital.

The remainder of the paper is structured as follows. Section 2 provides a brief review of the literature, highlighting our contribution to this area. Section 3 describes our sample and variable construction. Section 4 specifies our regression models and describes our estimation methods. Section 5 presents the results from our empirical analysis and finally Section 6 concludes.

2. Literature Related research in this area examines the link between firm’s corporate disclosure and its cost of equity. Studies such as Healy et al. (1999), Botosan (1997) and Botosan and Plumlee (2002), show that a reduction in information asymmetry between managers and shareholders, leads to a reduction in the cost of equity capital. Sengupta (1998) shows that the same relationship holds for the cost of debt. These studies however, only examine the disclosure dimension of corporate governance.

Concurrent work by Chen et al. (2003) and Ashbaugh et al. (2005), examine whether other corporate governance mechanism, apart from disclosure, have an impact on a company’s cost of equity. Chen et al. (2003) analyse this issue in the context of emerging markets. They examine firms from nine emerging Asian economies and find that disclosure and non-disclosure governance mechanisms such as board independence and minority shareholder protection, have a significant negative impact on a firms cost of equity capital. Similarly, Ashbaugh et al. (2005) find a negative relation between firm-level governance attributes and the cost of equity for US firms

4

from 1996 to 2002. Bhoraj and Sengupta (2003) supplement the above results by showing that similar firm-level governance mechanisms affect the cost of debt in the same manner. In comparison, our work examines this issue for Australian firms as well as looking at the overall cost of capital, rather than just the cost of equity.

This study is also related to the stream of literature that addresses the link between performance and mechanisms to control agency problems (McConnell and Servaes, 1995; Holderness et al., 1999; Agrawal and Knober, 1996; Demsetz and Villalonga, 2001; Himmelberg et al., 1999). In general, the empirical results from these studies are inconclusive. Given the ambiguity of research in this area, we provide an alternative approach to investigating this issue by examining variations in the cost of capital. Although the cost of capital is primarily a risk measure, it is also related to firm value. For example, a reduction in the cost of capital caused by strengthening a firm’s governance implicitly increases a firm’s market value. Chen et al. (2003) point out that existing literature in this area (Black et al., 2003; Claessens et al., 2003; Gompers et al., 2003; La Porta et al., 2002) assumes that governance affects firm valuation by increasing expected cash flow, since less cash flow is diverted away from shareholders. The idea that governance can enhance firm value through reducing the cost of capital, however, is not explicitly examined in these studies. The results of Chen et al. (2003) suggest that governance mechanisms do enhance firm value in this manner. The full extent of this relation, however, can be more accurately tested by considering a firm’s overall cost of capital (the cost of both its debt and equity). For firms that have a significant degree of leverage, capturing a reduction in the cost of equity will not reflect the full degree of an increase in firm value.

5

The use of the cost of capital to measure value has additional advantages over Tobin’s Q, which is widely used in the corporate finance literature. First, the measurement of Q is subject to accounting treatment of balance sheet items. Second, Q also reflects a firm’s growth opportunities. A change in a firm’s Q over time may simply reflect changes to the valuation of future growth opportunities which arise in part from factors exogenous to managerial decisions, such as economic and industry conditions. The cost of capital on the other hand reflects the required rate of return to capital, which is based on the current risk of the firm’s operations. The cost of capital is able to react more accurately to year to year changes to a firm’s governance environments without being influenced by exogenous factors that affect future growth and profitability. This is particularly important, given the findings of Himmelberg et al. (1999) and subsequent comments by Zhou (2001), which show that in a fixed effects estimation framework, year-to-year within-firm changes in firm value (as measured by Q) may be too noisy to detect the effects of typically small year-to-year changes in governance measures.

3. Data 3.1. Sample Selection Our data set initially comprises of the largest 150 Australian firms by market capitalisation. We delete listed financial and utility companies from the sample given their unique characteristics. We also delete firms for which we can not obtain a full set of variables described below. Our final sample comprises 136 firms and our period of investigation spans from 1994 to 2003. The ten-year window allows for

6

considerable variation in firm-level governance factors, which is important given the typically slow changes in these factors through time. The total number of firm-year observations is 861.

3.2. Firm-Level Corporate Governance Variables For each firm we collect information on three key governance mechanisms, (i) board independence and size, (ii) the extent of insider shareholdings and (iii) the extent of outsider shareholdings.

The board of directors’ role is to provide independent oversight of management and hold management accountable to shareholders for its actions. The fiduciary duty of the board of directors can be undermined if directors become allied with managers rather than protecting the interests of shareholders. In this sense, the lack of board independence from management is a governance risk that can materialize into reduced shareholder wealth. Previous studies examining the link between board structure and firm performance are inconclusive (Hermalin and Weishbach, 1991; Bhagat and Black, 2002; Brown and Caylor, 2004 and Agrawal and Knober, 1996). Therefore, our examination of the effect of board structure on the cost of capital provides an additional avenue to gain some insight into this issue. We measure board independence as the number of independent non-executive directors over the total number of directors (BoardIndep). We classify directors as non independent if they were current or ex employees, had business dealings with the firm, or were related (by family) to executive directors.

7

Board size is considered to be an independent governance mechanism (Jensen, 1993). The monitoring role of the board has been extensively studied and the general consensus is that smaller boards are more effective at monitoring and are related to higher firm value, (Jensen, 1993; Yermack, 1996; Eisenberg et al., 1998; Mak and Kusnadi, 2005). Smaller groups are more cohesive, more productive, and can monitor the firm more effectively. Larger groups are fraught with problems such as social loafing and higher coordination costs and hence are not good monitors. We measure board size as the natural logarithm of the total number of directors on the board. (LogBOARDSIZE). The information needed to construct board independence and size is hand collected from the firm’s annual report in the Connect 4 database.

We also hand collect shareholder information for each of the firms to construct ownership structure variables related to corporate governance. The first of these is the proportion of a firm’s stock that is held by corporate insiders, (INSIDER). The impact of insider ownership on firm value is actually non-monotonous. On the one hand, when managerial compensation is sensitive to firm performance, managers are more likely to pursue value maximizing strategies (the incentive effect). On the other hand, excessive insider ownership may insulate managers from outside shareholder monitoring, and managers may also begin to pursue risk reduction strategies to protect their large undiversified shareholding. Thus, very large controlling shareholdings by insiders can adversely affect firm value (the entrenchment effect). We use the square of insider ownership (INSIDER2) to control for the nonlinear (inverted U-shape) relationship between value and managerial ownership, (McConnell and Servaes, 1990).

8

The second ownership variable is the percentage ownership belonging to institutional block shareholders (INSTBLOCK), where a block holding is defined as a holding which is equal to or greater than 5% of total ownership. Jensen (1993) and Shleifer and Vishy (1997) argue that block shareholders are important for effective corporate governance since they can exercise their voting power to curb value destroying behaviour by management. The relatively large size of their shareholdings also provides a greater incentive to monitor than those of dispersed small shareholders.

Institutional block shareholders such as banks, superannuation (pension) funds and mutual funds are a unique class of blockholders since they are likely to be independent of management and have the ability to intervene or place pressure on management to protect their minority interest. Further, Cremers and Nair (2005) argue that pension funds face fewer conflicts of interest than other institutional investors and they tend to be aggressive shareholder activists that are effective in monitoring the activities of management. To the extent blockholders and activist institutional investors provide effective monitoring of management that reduces opportunistic behaviour, all shareholders benefit leading to a reduction of agency risk and a lower cost of capital.

Finally, we examine whether this argument can be applied to other block shareholders that are not financial institutions. For our sample firms, these shareholders are mostly parent and associate companies. We measure their total percentage ownership and label this variable NONINSTBLOCK. The effect of this variable on corporate value and performance is ambiguous since these block shareholders are also capable of colluding with each other and with insiders to expropriate minorities.

9

Table 1 presents the overall descriptive statistics on these variables. Australian firms have notably higher insider ownership concentration than their US counterparts. The average of the variable INSIDER for our sample is about 12%, while the equivalent statistic for US firms documented in Ashbaugh et al. (2005) is around 6%. The larger insider shareholdings appear to correspond with less outside influence for Australian firms. The average proportion of independent directors in our sample is about 56%, compared to 66% for US firms as reported by Ashbaugh et al. (2005). Institutional blockholders own an average of 15% of shares. For US firms, Ashbaugh et al. (2005) report that the average percentage ownership of all institutional investors is about 65% and Cremers and Nair (2005) document that the largest institutional blockholder alone already owns about 8% of issued shares. These statistics reflect the fact that unlike those in the US and UK, many of the largest Australian firms are not widely held. This implies that the incentive (or entrenchment) effect of insider ownership may be highly observable for Australian firms.

[INSERT TABLE 1 HERE]

The statistics in Panel B of Table 1 describe within-firm changes in these governance variables through time. One potential issue with using the fixed-effects regression methodology is the lack of substantial variations in the explanatory variables, leading to insignificant coefficient estimates. Zhou (2001), documents that the yearly absolute change in CEO ownership is less than 10% for about 50% of US firms. In contrast, Australian firms appear to display a larger extent of changes in ownership and board structure from one year to another, and thus alleviates this issue. For example, the yearly absolute change of the INSIDER variable is 1.33 percentage points for our sample firms, equivalent to about 11% of the average insider ownership. Zhou (2001)

10

also reports that among closely held US firms (i.e. those with CEO own more than10% of issued shares), only 26% display a yearly absolute CEO ownership change of 10% or above. The equivalent statistic for our sample firms is much higher at 40%. Furthermore, the institutional blockholders and board independence statistics also display large changes from one year to another, with an average absolute change of 5.54 and 3.7 percentage points per year. The latter figure is equivalent to the addition (or reduction) of one independent director in about every two years. Overall, these statistics illustrate that there is a significant degree of year-to-year of variation in our governance variables, which permits a richer analysis of the effect of governance on performance.

3.3. Firm-Level Characteristics Firm-level control variables that potentially influence the cost of capital and firm value are collected from the FinAnalysis database provided by Aspect Financial. We employ the ratio of capital expenditures to capital stock to control for the scope of discretionary spending in growth firms (CAPEX/TA), the log of total assets to control for firm size (LogTA) and the ratio of tangible assets to total assets to control for asset tangibility (TANA/TA). These variables control for the firm-level agency environment and information asymmetry that could intervene in the relationship between governance and the cost of capital. The ratio of total debt to assets (TL/TA) is used to control for the effect of leverage on the cost of capital and the book value to market value of equity ratio (BM) is used to control for the effect of a firm’s growth prospects on its cost of capital. Similar control variables are used in other studies that examine the effect of governance on firm value. (Ashbaugh et al., 2005). The standard

11

deviation (SD) of monthly stock returns calculated over a rolling 60-month window is used to control for total risk in the regressions involving Q. However we omit this variable in the regressions involving the cost of capital since this measure is based on a firm’s risk.

3.4. Cost of Capital To measure the cost of capital we obtain the estimated weighted average cost of capital (WACC) from Stern Stewart & Co. This measure of a firm’s cost of capital is used to calculate the Stern Stewart EVA measure which is widely accepted performance benchmark. The weighted average cost of capital is calculated as:

WACC = (D/EV x (1-t) x Kd) + (E/EV x Ke)

(1)

where D/EV = Debt to Enterprise Value ratio which is established using a three year trailing average of D/EV levels. E/EV is the ratio of the firm’s equity to its enterprise value. t is the income tax rate for companies. Kd is the cost of debt. As debt is not listed for most Australian companies, the yield to maturity is difficult to estimate. The method therefore makes the simplifying assumption that all debt is BBB rated and uses the BBB spread above the risk free to estimate the pre-tax cost of debt. Ke is the cost of equity capital, calculated using the Capital Asset Pricing Model as follows:

Ke = Rf + ERP x ß

(2)

Rf is the risk free rate calculated using the average yield on 10 year Australian government bonds. ERP is the equity risk premium and is assumed to be 6%. BETA

12

(ß) is calculated using daily prices from the Australian Graduate School of Management’s CRIF Database.

Given that we rely on external estimates of the cost of capital, we conduct validation regressions to ensure that the estimates are sufficient proxies. We regress WACC estimates on variables that are known to influence a firm’s expected returns. These include a beta measure based on our own calculation (BETA), leverage (TL/TA), size (LogTA) and the book-to-market ratio (BM). The results are reported in Table 2, and indicate that our WACC estimates are strongly related to factors that should influence a firms cost of capital. In particular, BETA is able to explain 33 percent of the variation in our WACC estimates. The proportion of debt (TL/TA) is negatively related to the cost of capital, illustrating the effect of leverage in decreasing a firms cost of capital and increasing returns to shareholders. Specification 5 in Table 2 shows that in total, our selected factors are able to explain 49.3 percent of the variation of our WACC estimates, providing a reasonable degree of confidence in our cost of capital estimates.

[INSERT TABLE 2 HERE]

3.4. Firm Value Measures In order to provide comparisons between our results and previous studies, we replicate regressions that test the relation between firm value and governance mechanisms. We use Tobin’s Q (Q), defined as the market value of equity plus the book value of debt over total assets as our measure of firm value. Table 1 reveals the considerable standard deviation in Q, underscoring our argument that this measure is

13

influenced by a wide variety of factors that make it difficult to detect the true impact of governance changes. Our measure of cost of capital on the other hand appears to show considerably less variation compared to Q, with the standard deviation (0.03) being less than one-third of the mean (0.10).

4. Methodology Following Himmelberg et al. (1999), the relation between corporate governance factors is estimated using a fixed-effects panel regression. This method accounts for any potential endogeneity of our governance measures (particularly inside ownership, and institutional block shareholdings) by controlling for potential unobserved firmspecific factors that could be driving both governance mechanisms and performance.

The model is specified as follows:

WACCit = β 0 + β1BOARDINDEP + β 2 LogBOARDSIZE + β 2 INSIDER + β 2 INSIDER 2 K

+ β 3 INSTBLOCK + β 4 NONINSTBLO CK + ∑ δ k xitk + λi + ε it

(3)

k =1

where : K

∑δ k =1

k k it

x denotes our set of control variables and λi denotes firm-level fixed effects.

BOARDINDEP is the percentage of independent non-executive directors on the board, LogBOARDSIZE is the logarithm of the number of directors on the board, INSIDER is

the percentage executive directors shareholding, INSTBLOCK is the percentage institutional shareholding and NONINSTBLOCK is the percentage ownership of noninstitutional block shareholders (not including insiders)

14

The firm-level fixed effects model assigns a unique identification variable to each firm in the sample denoted by λi . By including this variable in our estimation, the model controls for any firm-specific factors that could be driving both governance mechanisms and performance. For example a particular firm may have outside owners that have access to superior monitoring technology and thus require managers to own less stock to align their incentives. Access to superior monitoring technology may also reduce the rate of return required by stock and bond holders, and therefore this unobserved factor could lead to an incorrect finding that lower insider ownership is associated with a lower cost of capital, when in fact the existence of a firm-specific factor (superior monitoring technology of the owners) is driving this association. It is important to note that the fixed-effects model co-efficients relate only to within-firm changes over time and do not take into account any variation across firms.

5. Results 5.1. Firm Value Regressions In order to provide insight into the significance of our approach vis-à-vis past results, we first replicate the fixed effects OLS regressions estimated by Himmelberg et al. (1999) which examined the effect of governance mechanisms on firm value. The results in Table 3 confirm the results in Himmelberg et al. (1999). There is no relationship between any of the governance variables and Tobin’s Q. Even though fixed effects estimation is considered a suitable approach to deal with the endogeneity problem, Zhou (2001) explains that by restricting the scope of estimation to withinfirm changes over time, a relationship between managerial ownership and performance cannot be detected even if one exists. 15

It is important to note that by using the cost of capital to infer performance, we use a dependant variable that has significantly less variation than Q. Moreover, as the ownership-related governance variables display significant within-firm variation over time, this gives us a much better chance of detecting a relationship, whilst also making use of the ability of fixed effects estimation to account for unobserved firm heterogeneity.

[INSERT TABLE 3 HERE]

5.2. Cost of Capital Regressions The relation between corporate governance factors and the cost of capital is also estimated using a fixed-effects panel regression. Table 4 reports the estimation results for the cost of capital measure regressed on combinations of the governance variables as well as the set of controls described in section 3.2. The results show that corporate governance variables play a significant role in explaining the variations in a firm’s cost of capital. Institutional block holdings (INSTBLOCK) is significantly negatively related to the cost of capital across all specifications of the model, suggesting that the higher institutional block holdings, the lower is the cost of capital. The presence of financial institutions as block shareholders reduces the risks associated with the provision of capital as they ensure that cash flows are not diverted away from shareholders, and that capital is used optimally to maximise shareholder wealth. This result is consistent with Cremers and Nair (2005) and Ashbaugh et al. (2005).

16

Further, insider ownership (INSIDER) has a significant negative relationship with cost of capital across all specifications of the model, suggesting that the higher insider ownership, the lower is the cost of capital. This result is consistent with the results of Ashbaugh et al. (2005) for the cost of equity. The (INSIDER2) term however, although not significant, is the opposite sign of INSIDER, indicating the non-monotonic relation between insider ownership and the cost of capital. If insider ownership is too high, the risk of managerial entrenchment rises thereby increasing the cost of capital.

The results also indicate that an increase in board independence (BOARDINDEP) significantly decrease a firm’s cost of capital. The results are consistent with Ashbaugh et al. (2005) and Chen et al. (2003) for the cost of equity. Further, a smaller number of directors on the board (logBOARDSIZE) also significantly decrease a firm’s cost of capital. This suggests that the presence of a small and focused board whose monitoring incentives are aligned can significantly increase a firm’s valuation. The result is consistent with earlier studies that find that smaller boards are more effective at monitoring and are related to higher firm value, (Jensen, 1993; Yermack, 1996; Eisenberg et al., 1998; Mak and Kusnadi, 2005).

The results for NONINSTBLOCK are the weakest of all the governance variables. This result is consistent with our expectation that some block holders are also capable of colluding with management to expropriate other minorities and thus their presence could act to increase the cost of capital. Institutional shareholdings on the other hand are much less likely to be perceived in this way.

[INSERT TABLE 4 HERE]

17

Of the control variables, the size variable (LogTA) is significant throughout all specifications. This indicates that large firms are likely to have a lower cost of capital. From a governance perspective, large firms may be more transparent and thus easier to monitor, leading investors to demand lower returns. Investment cash flow over total assets (CAPEX/TA) is positively related to the cost of capital, since high investment or growth firms are more likely to have higher returns demanded from them. Leverage (TL/TA) has a negative and significant relationship with the cost of capital, illustrating that firms that are able to absorb more debt are able to take advantage of the debt tax shield and reduce their cost of capital.

6. Conclusion

We use a sample of large Australian firms from 1994 to 2003 to examine the valuecreation role of corporate governance. We employ a fixed-effects regression model to control for the endogeneity of our governance measures and use the cost of capital as an alternative value measurement. We show that variation in firm-level corporate governance mechanisms plays an important role in explaining variations in firms’ cost of capital. Our empirical results show that greater insider ownership, the presence of institutional blockholders, and smaller and more independent boards all serve to reduce the perceived risk and level of information asymmetry of a firm, thereby leading investors to demand lower rates of return on capital provided. This highlights the important role that corporate governance plays in creating value for shareholders. Given the inconclusiveness of existing literature that uses Tobin’s Q to measure firm

18

value, this research provides an alternative and potentially more suitable way to investigate the impact of corporate governance on firm value.

19

References

Agrawal, A., Knoeber, C.R., 1996,. Firm Performance and Mechanisms to Control Agency Problems between Managers and Shareholders, Journal of Financial and Quantitative Analysis 31, 377.397.

Ashbaugh, H., Collins, D.W., LaFond, R., 2005. Corporate Governance and the Cost of Equity Capital, Working Paper, University of Iowa.

Bhagat, S., Black, B., 2002. Board Independence and Long-Term Performance, Journal of Corporation Law 27, 231.273.

Bhojraj, S., Sengupta, P., 2003. Effect of Corporate Governance on Bond Ratings and Yields: The Role of Institutional Investors and Outside Directors, The Journal of Business 76, 455.475.

Black, B., Jang, H., Kim, W., 2003. Does corporate governance affect firm’s market values? Evidence from Korea, Working paper, Stanford Law School.

Brown, L., Caylor, M., 2004. Corporate Governance and Firm Performance, Working paper, Georgia State University.

Botosan, C., 1997, Disclosure level on the cost of equity capital, Accounting Review 72, 323.349.

20

Botosan, C., Plumlee, M., 2002. A re-examination of disclosure level and the expected cost of equity capital, Journal of Accounting Research 40, 21.40.

Chen, K.C.W., Chen, Z., Wei, K.C.J., 2003. Disclosure, Corporate Governance, and the Cost of Equity Capital: Evidence from Asia’s Emerging Markets, Working Paper, Hong Kong University of Science and Technology.

Claessens, S., Djankov, S., Fan, J.P.H., Lang, L.H.P., 2002. Disentangling the incentive and entrenchment effects of large shareholdings, Journal of Finance 57, 2741.2771.

Cremers, M., Nair, V. B., 2004. Governance Mechanisms and Equity Prices, Journal of Finance 60 , 2859.2894.

Demsetz, H., Lehn, K., 1985. The structure of corporate ownership: causes and consequences, Journal of Political Economy 93, 1155.1177.

Demsetz, H., Villalonga, B., 2001. Ownership structure and corporate performance, Journal of Corporate Finance 7, 209.233.

Eisenberg, T., Sundgren, S., Wells, M., 1998, Larger Board Size and Decreasing Firm Value in Small Firms, Journal of Financial Economics 48, 35.54.

Gompers, P. Ishii, J., Metrick, A., 2003. Corporate Governance and Equity Prices, Quarterly Journal of Economics 118, 107.155.

21

Healy, P., Hutton, A., Palepu, K., 1999. Stock performance and intermediation changes surrounding increases in disclosure, Contemporary Accounting Research 16, 485.520.

Hermalin, B., Weisbach, M., 1991. The Effect of Board Composition and Direct Incentives on Firm Performance, Financial Management 21, 101.112.

Himmelberg, C., Hubbard, R.G., Palia, D., 1999,. Understanding the determinants of managerial ownership and the link between ownership and performance, Journal of Financial Economics 53, 353–.384.

Holderness, C., Kroszner, R., Sheehan, D., 1999. Were the good old days that good? Evolution of managerial stock ownership and corporate governance since the great depression, Journal of Finance 54, 435–469.

Jensen, M., 1993. The Modern Industrial Revolution, Exit, and the Failure of Internal Control Systems, Journal of Finance 48, 831.880.

La Porta, R., Lopez-De-Silanes, F., Shleifer, A., 2002. Investor protection and corporate valuation, Journal of Finance 57, 1147.1170.

Mak, Y., Kusnadi, Y., 2005. Size really matters: further evidence on the negative relationship between board size and firm value, Pacific Basin Finance Journal 13, 301.318.

22

McConnell , J., Servaes, H., 1990. Additional Evidence on Equity Ownership and Corporate Value, Journal of Financial Economics 27, 595.612.

McConnell , J., Servaes, H., 1995. Equity Ownership and the Two Faces of Debt, Journal of Financial Economics 39, 1995.

Sengupta, P., 1998. Corporate disclosure quality and the cost of debt, Accounting Review 73, 459.474.

Shleifer, A., Vishny, R., 1997. A Survey of Corporate Governance, Journal of Finance 52, 737.783.

Yermack, D., 1996. Higher Valuation of Companies with a Small Board of Directors, Journal of Financial Economics 40, 185.212.

Zhou, X., 2001. Understanding the determinants of managerial ownership and the link between ownership and performance: comment, Journal of Financial Economics, 62, 559.571.

23

Table 1. Descriptive Statistics of the Sample Firms the Period of 1994 to 2003 Panel A provides descriptive statistics for our dependent and independent variables described in section 3. Statistics are calculated based on pooled data across all firms and all years. Panel B provides statistics on the average year-to-year change in each of our governance variables. Changes are calculated as absolute values such that subsequent changes in the opposite direction are not nullified.

Mean

WACC

Standard Deviation

Min

Median

Max

Skewness

PANEL A: Pooled-Data Descriptive Statistics 0.10 0.03 0.05 0.10

0.22

0.80

Q

1.83

1.64

0.50

1.37

17.05

5.12

CAPEX / TA (%)

8.57

13.42

-95.92

6.26

91.60

0.69

TL/TA (%)

50.13

17.42

0.83

51.48

186.94

0.48

TanA/TA (%)

84.22

20.97

10.41

93.53

100.00

-1.49

TA ($ billion)

2.76

7.67

0.01

0.74

101.51

7.03

BM

0.65

0.47

0.03

0.56

4.25

2.22

53.96

24.61

0.00

57.14

100.00

-0.51

7.83

7.50

3.00

7.41

20.00

1.02

INSIDER

11.65

19.99

0.00

0.00

82.54

1.55

INSTBLOCK

14.51

14.30

0.00

11.27

96.77

0.92

NONINSTBLOCK (%)

11.74

21.00

0.00

0.00

96.00

1.95

BOARDINDEP (%) BOARDSIZE

PANEL B: Year-to-Year Governance Changes (Absolute Value) Statistics ΔBOARDINDEP (%)

3.71

6.53

0.00

0.79

66.67

3.71

ΔBOARDSIZE

0.62

0.85

0.00

0.00

6.00

1.55

ΔINSIDER (%)

1.33

4.94

0.00

0.00

70.69

7.43

ΔINSTBLOCK (%)

5.54

7.26

0.00

3.44

62.55

2.65

ΔNONINSTBLOCK (%)

2.41

7.86

0.00

0.00

92.00

6.07

WACC = Weighted average cost of capital Q = Market value of equity plus book value of debt divided by book value of total assets EVA / TA = Economic value added divided by total assets CAPEX / TA = The ratio of capital expenditures to total assets TL / TA = Total liabilities over total assets TANA / TA = Tangible long term assets (property, plant and equipment) over total assets TA = Book value of total assets BM = Book Value to Market Value Ratio INSTBLOCK = Percentage ownership of institutional block shareholders NONINSTBLOCK = Percentage ownership of non-institutional block shareholders (not including insiders) INSIDER= Percentage ownership of insider block shareholders BOARDINDEP = Proportion of directors who are independent non-executives BOARDSIZE = The number of directors

Table 2. 24

Cross-Sectional Validation of the Stern Stuart Cost of Capital Measure The table provides the cross-sectional OLS regression estimates for Stern Stewart’s cost of capital measure regressed on BETA, size (LogTA), market-to-book (MB) and Leverage (TL/TA). All variables are averaged across the sample period such that each firm is represented by a single observation in the regression (136 observations). White-adjusted standard errors are reported below each of the coefficients.

Dependent Variable: Average Cost of Capital 1 Intercept BETA

2 a

0.084

0.003

3 a

4

5

0.131

a

0.113

a

0.129

0.103a

0.028

0.0039

0.010

0.022

a

0.018

0.016a

0.003

0.0026

LogTA

-0.001

0.001

0.001

0.001

MB

a

-0.013a

-0.016 0.005

TL/TA

Adjusted R

0.004 -0.052

2

Observations

b

-0.054a

0.021

0.010

0.333

0.000

0.074

0.135

0.493

136

136

136

136

136

a b

, , and c indicate significance at the 1%, 5%, and 10% levels

BETA = Beta of individual firm calculated using monthly stock returns LogTA = Natural logarithm of total assets MB = Market value to book value Ratio TL / TA = Total liabilities over total assets

25

Table 3 Fixed-effects Regression between Tobin’s Q and Governance Variables The table provides the results for the fixed-effects regressions of our dependent variables on our governance variables and controls described in section 3. The dependant variables is Tobin’s Q. Whiteadjusted standard errors are reported below each of the coefficients

Dependent Variable: Q

LogTA CAPEX / TA TL / TA TANA / TA SD

1

2

3

-0.017 (0.116) -0.610 (0.506) -1.164a (0.448) 2.139a (0.695) -0.339 (1.924)

0.029 (0.114) -0.572 (0.508) -1.263a (0.453) 2.206a (0.696) 0.029 (0.114) 0.209 (0.337) -0.401 (0.244)

0.003 (0.121) -0.578 (0.511) -1.159a (0.448) 2.170a (0.697) -0.507 (1.950) 0.230 (0.345) -0.392 (0.249) -0.775 (1.161) 0.067 (1.922) -0.075 (0.289) 0.040 (0.291)

0.645

0.645

BOARDINDEP LogBOARDSIZE INSIDER INSIDER2 INSTBLOCK NONINSTBLOCK Adjusted R2 a b

-0.888 (1.127) 0.155 (1.866) -0.018 (0.282) -0.062 (0.286) 0.643

c

, , and indicate significance at the 1%, 5%, and 10% levels

LogTA = Natural logarithm of total assets CAPEX / TA = The ratio of capital expenditures to total assets TL / TA = Total liabilities over total assets TANA / TA = Tangible long term assets (property, plant and equipment) over total assets SD = Standard deviation of weekly stock returns for each calendar year BOARDINDEP = Proportion of directors who are independent non-executives LogBOARDSIZE = Logarithm of the number of directors INSIDER= Percentage ownership of insider block shareholders INSTBLOCK = Percentage ownership of institutional block shareholders NONINSTBLOCK = Percentage ownership of non-institutional block shareholders (not including insiders)

26

Table 4 Fixed-effects Regression between Cost of Capital and Governance Variables The table provides the results for the fixed effects regressions of the cost of capital on our governance variables and controls described in section 3. The dependant variable is the Stern Stewart cost of capital measure. White adjusted-standard errors are reported below each of the coefficients.

Dependant Variable: WACC LogTA CAPEX / TA TL / TA TANA / TA BM BOARDINDEP LogBOARDSIZE

1 -0.461a (0.155) 1.051b (0.460) -2.525a (0.602) -0.150 (1.116) -0.089 (0.211) -1.774a (0.671) 1.039b (0.482)

2 -0.552a (0.152) 1.111b (0.471) -2.245a (0.613) -0.205 (1.069) -0.091 (0.214)

INSIDER

-6.525b (3.248) 6.436 (4.537) -1.702b (0.812) 0.018 (0.794) 0.614

INSIDER2 INSTBLOCK NONINSTBLOCK Adjusted R2

0.616

3 -0.563a (0.152) 0.922b (0.460) -2.244a (0.611) -0.245 (1.068) -0.095 (0.207) -1.863a (0.697) 1.001b (0.477) -7.332b (3.200) 7.350 (4.540) -1.529c (0.796) -0.586 (0.772) 0.623

a b

, , and c indicate significance at the 1%, 5%, and 10% levels

LogTA = Natural logarithm of total assets CAPEX/TA = The ratio of capital expenditures to total assets TL / TA = Total liabilities over total assets TANA/ TA = Tangible long term assets (property, plant and equipment) over total assets BM = Book Value to Market Value Ratio BOARDINDEP = Proportion of directors who are independent non-executives LogBOARDSIZE = Logarithm of the number of directors INSIDER= Percentage ownership of insider block shareholders INSTBLOCK = Percentage ownership of institutional block shareholders NONINSTBLOCK = Percentage ownership of non-institutional block shareholders (not including insiders)

27