Corporate Governance and the Cost of Capital Review of the Empirical Literature.docx

《Corporate Governance and the Cost of Capital Review of the Empirical Literature.docx》由会员分享，可在线阅读，更多相关《Corporate Governance and the Cost of Capital Review of the Empirical Literature.docx（25页珍藏版）》请在淘文阁 - 分享文档赚钱的网站上搜索。

1、CORPORATE GOVERNANCE AND THE COST OF CAPITAL: REVIEW OFTHE EMPIRICAL LITERATUREZulkufly RamlyFaculty of Business and AccountancyUniversity of Malaya, MalaysiaE-mail: zulramly um. edu. myABSTRACTCorporate governance encompasses a broad spectrum of mechanisms intended to mitigate agency risk by increa

2、sing the monitoring of managements5 actions, limiting managers opportunistic behaviour, and improving the quality of firms9 information flows. A torrent of literature explains that corporate governance (CG) mechanisms such as quality of information disclosure, ownership structure, independent direct

3、ors, audit committee, institutional shareholders are able to contribute toward improving firms performance. Indeed, robust CG is expected to contribute to the overall value creation process (Shleifer and Vishny, 1997). The other dimension of value creation is the reduction in the cost of capital (CO

4、C) raised by firms. Theoretically and empirically to some extent, good CG will lead to lower firm risk and subsequently to a lower cost of capital. Firms that are well-managed in terms of the existence of robust monitoring devices as well as the provision of quality financial reporting and protectio

5、n of stakeholders5 well-being will be able to limit the exercise of power of corporate managers and prudently allocate resources, which in turn enjoy lower risk than other firms. It follows that these firms should have access to cheaper source of capital, either in the form of equity or debt or both

6、, than other firms. This paper aims to provide a critical review of the empirical literature on the effect of CG on costs of both equity and debt capitals.Keywords: Corporate Governance, Agency Theory, Cost of Equity Capital, Cost of Debt.INTRODUCTIONIn modern public corporations suppliers of financ

7、e or the owners do not have full control over the spending of their money and have limited influence over decision making process. The owners surrender the control to professional controllers or managers who exert immense control over the resources of a firm. In essence in public firms the ownership

8、 is separated from control. The separation of ownership and control (Berle and Means, 1932) leads to conflicts of interest between managers and owners. Conflicts of interest between managers and owners arise when managers engage in activities that are not in line with the objective of maximizing sha

9、reholders9 wealth. Owners of a firm simply want the value of their shares to be as high as possible and entrust upon managers to undertake activities and investments that support this aspiration. However, managersCost of equity capital estimationThis section begins with the discussion on the common

10、pitfalls in the approaches used in estimating cost of equity. Basically, the drawbacks lie on the difficulty to estimate risk loadings, risk factor premiums and problem with appropriate model selection. Next, a review on the mechanics of two estimation models namely the dividend discount model or th

11、e residual income model will be provided.The cost of equity capital is the discount rate the market applies to a firms future cash flows to determine current share price. It is an important factor in deciding an investment choice and evaluating an existing investment. However, based on the past lite

12、rature, researchers have yet to reach a consensus on one single well-accepted method in estimating the elements in cost of equity capital estimation in academic research (see Botosan and Plumlee, 2005; Gode and Mohanram, 2003; Guay, Kothari and Shu; Easton and Mohanram, 2005).Fama and French (1997)

13、studied two methods of cost of equity estimation namely the Capital Asset Pricing Model (CAPM) and the Fama and French three-factor pricing model and discovered that the estimates of the cost of equity for industries are imprecise. The typical standard errors in estimating the cost of equity is more

14、 than 3 percent for both models. They identified three potential limitations of these two approaches that could possibly explain the large standard error, which leads to the imprecise estimates of the industry cost of equity. The first limitation relates to the uncertainty about the accurate estimat

15、e of factor risk premiums. The use of historical market premium to estimate the expected premium resulted in a large standard error of estimates. The second limitation is due to the variation in risk factor loadings for industries through time whereas the models call for constant loadings for the es

16、timate to be accurate. The third limitation is due to the difficulty in selecting the right model. Although both models utilise the same estimate of the market risk premium, the estimates vary by a relatively large fraction. Ultimately, when the estimate of industry cost of equity is imprecise it wi

17、ll be problematic to derive an accurate estimate for individual firms and investment projects because it can be expected that the standard error of firms cost of equity to be even larger.The asset pricing theory calls for the use of expected returns method in estimating the cost of equity. Based on

18、the Capital Asset Pricing Model (CAPM), a discount rate is used to estimate the present value of expected dividends. A discount rate is termed as the sum of the equity risk premium plus risk-free-rate. Equity risk premium is not directly observable, thus the expected returns are used to infer it. Th

19、e estimate is either based on ex post from realised return or ex ante (expected) from the current price and expectations of future dividends. Prior empirical research mainly utilised the average ex post (realised) realised returns because the ex ante returns are not observable. This practice is driv

20、en by the assumption that when the market is efficient, risk is appropriately priced, which makes the average realised return an unbiased estimator of the unobservable ex-ante (expected) returns (Gebhardt, Lee and Swaminathan, 2001). One major limitation of the estimates of cost of equity derived fr

21、om the average realised returns lies on the difficulty in establishing a significant association between the returns and market beta, which is awidely accepted measure of risk and known to have a considerable influence on a Erms cost of capital. Instead, previous studies discovered that the realised

22、 returns had a significant association with other variables such as book-to-market and size, which have little relation with average returns (Fama and French, 1992; Elton, 1999). This finding is in contrary to the main premise of the asset pricing theory that asserts the power of market beta in expl

23、aining the cross-section of average returns. The key advantage of using ex ante cost of equity capital to measure cost of equity lies on its ability to make an explicit control for cash flows and growth potential (Hail and Leuz, 2006) and may provide a better measure for the expected return (Pastor,

24、 Sinha and Swaminathan, 2021).The dividend discount model (DDM) uses the ex ante approach in estimating the cost of equity. It is attractive and forward-looking approach because it infers the risk premium from the current share price and future expected dividends (Gode and Mohanram, 2003). The DDM e

25、quates current share price to a finite series of expected future cash flows and a terminal value, discounted using cost the cost of equity capital. Two key assumptions are made in the model: (1) a pattern of dividend payout ratio and (2) the terminal value at the end of the forecast horizon. The bul

26、k of the expected future cash flows lie in the terminal value. However, market expectations and future dividends are not observable necessitating the use of analysts earnings forecasts to proxy fbr market expectations. Hence, an accurate inference of the cost of equity depends on the ability to reco

27、gnize the market9s terminal value forecast (Botosan and Plumlee, 2005).The next estimation model is known as the residual income valuation (RIV), which is derived from the DDM. The RIV does not rely on average realized returns in estimating the implied cost of equity. The implied cost of equity is d

28、efined as the internal rate of return (IRR) that equates current share price to the present value of all future cash flows to common shareholders. This approach requires two procedures: (1) forecasting of cash flows until a terminal period and (2) projecting a terminal value that is the intrinsic va

29、lue of a firm based on the residual income earned beyond the explicit forecasting period. This approach explicitly uses analysts5 forecasts to proxy fbr the market expectation of a firms earnings. In essence, the RIV is derived from the DDM but provides better insight on the role of economic profits

30、 on share valuation and expresses firm value in terms of accounting numbers (Gebhardt, et al., 2001). The variations in the RIV approach are explained in the models of Gebhardt et al. (2001), Claus and Thomas (2001), Easton (2004) and Ohlson and Jauttner-Nauroth (the OJ Model) (2005). Appendix A sum

31、marises the salient features of each of the models.Measures of cost of debtThe cost of debt is basically the cost incurs when a firm obtains external financing from lenders or fund providers. The most common proxy for cost of debt used in prior studies is the yield spread (e.g. Ertugul and Hedge, 20

32、21; Duffie, 1998; Anderson, et aL, 2003; Anderson, Mansi and Reeb, 2004; Klock, Mansi and Maxwell, 2005). The yield spread is basically the weighted average debt yield to maturity in excess of the duration equivalent to treasury yield. This cost of debt measure is typically used in the fixed- income

33、 literature to compute the debt risk premium (Duffie, 1998).Prior studies had also utilised yield to maturity on the first debt issue of year t + 1 to proxy for cost of debt (see Sengupta, 1998; Blom and Schauten (2006). This yield to maturity represents the effective rate of interest that equates t

34、he present value of the principal and interest payments with the amount paid by the lender. Other measures of cost of debt include the average interest on a firms debt and the total interest cost to the firm on its first debt issue of year t + 1 (Sengupta, 1998; Piot & Missonier-Piera, 2007). Table

35、2 below summarises key measure of cost of debt utilised in prior research. Appendix B summarises the salient features of each of the cost of debt measures.Measure of the overall cost of capitalBased on the review of past literature it appears that only one research explored the impact of corporate g

36、overnance on a combination of equity and debt capital of a firm (see Pham, Suchard and Zein, 2007). They utilised the estimated weighted average cost of capital (WACC) to reflect the actual sources of capital of many firms. The WACC considers the fact that many firms obtain capital from both sources

37、 namely equity and debt. They obtained the WACC measure from Stern Stewart & Co, which is widely accepted performance benchmark. The WACC is a good proxy for a firms cost of debt if it is strongly related to factors that should influence a firms cost of capital (Pham, et al., 2007). The factors incl

38、ude beta, leverage, size and book-to-market ratio.CORPORATE GOVERNANCE AND THE COST OF CAPITALFrom a theoretical perspective, empirical investigation proposes that weak CG will lead to amplified agency risk, which may increase the uncertainty of future cash flows (Jensen and Meckling, 1976; Jensen,

39、1986 and Bhojraj and Sengupta, 2003). Summing up the prior empirical studies, CG has been found to have a linkage with the COC, with stronger governance leads to lower COC. Previous studies in this strand of research have utilised both specific CG variables and CG index as proxies for a firms govern

40、ance quality and examined its impact on measures of cost of capital. A CG index is a score card that measures the quality of a firms CG over several dimensions. It is imperative to note that prior studies mainly examine the effects of CG on COEC and COD separately.The following section reviews findi

41、ngs of prior research with regards to the relationship between CG and COEC.Corporate governance and the cost of equity capitalIn the U.S.A context, Ashbaugh, Collins, and LaFond, (2004) documented the effect of corporate governance on the COEC of U.S. firms by linking governance attributes to frms e

42、xpected returns, beta and realised returns. The governance attributes used in this research were related to (1) quality of firms financial information, (2) ownership structure, (3) shareholder rights, and (4) board structure. These mechanisms are intended to reduce moral hazard and adverse selection

43、 problems present in public listed companies. The authors used two measures of COEC: (1) Target method using the average firms expected return over its fiscal period as employed in Botosan & Plumlee (2002, 2005) and Francis, Khurana, & Pereira (2005), (2) using PEG ratio as developed in Easton (2004

44、). A composite CG score fbr each firm is constructed to capture a firms overallgovernance risk. Overall, they found that the governance attributes significantly affected firms COEC directly, as well as indirectly via systematic risk (p), since most of the CG attributes were significantly associated

45、with 0.Using a sample of 8,836 firm-year observations, Huang (2004) investigated the effect of firm-level variation in shareholder rights on the ex-ante COEC. In this study shareholder rights simply mean the ability of shareholders to remove managers. Weak shareholder rights indicates that poorly pe

46、rforming managers are able to entrench themselves, thus raising the COC. An alternative hypothesis posits that weak shareholder rights creates job security among managers, thus reduces managerial myopia and motivates them to allocate funds for beneficial long-term projects. This helps to reduce COC.

47、 Governance Score (G-score), which is adopted from Gompers, Ishii and Metric (2003) (later known as the GIM Index), represents the extent of shareholder rights. The five dimensions of the GIM Index include (1) tactics for delaying hostile bidders, (2) voting rights, (3) director/officer protection,

48、(4) other takeover defences and (5) state laws. Every provision of the GIM Index that restricts shareholder rights and increases managerial power are given one point. As such, high the G-score indicates weaker level of shareholder rights in a firm. The COEC estimate is based on the OJ Model abnormal

49、 earnings-based valuation model. Using both pooled and cross-sectional regression techniques, the results indicate that weak shareholder rights (higher G-score) are significantly associated with higher COEC. The study also found that there is a significant association between the change in G-score and the change in COEC. The results support the notion that weak shareholder rights leads to higher agency cost and the efficient market captures this effect into the COEC.Using the Gompers, et al. (2003) data from 1992 through 200