4.1 Summary statistics and correlations
Table 1 provides descriptive statistics for the samples used in our tests. The sample has a mean industry-adjusted Tobin’s Q of 0.49634. The minimum G-index is 1, while the maximum G-index is presented as 19, broadly consistent with the numbers shown by GIM (2003) for firms in the United States from period 1990 to 1998. It means that the assumption we made about the stability of changing in anti-takeover provisions is reasonable. The same situation is shown as E-index. The average total debt to total assets ratio is 0.21942, consistent with the values found by Rajan and Zingales (1995) for firms in developed countries. From Table 1 we can observe that the mean profitability, measured as EBIT divided by sales, is quite small. The estimated coefficient of profitability might be difficult to explain due to a great deal of negative EBIT.
Table 2 shows the Pearson correlation coefficients of all variables used. The strong relation between Tobin’s Q and the proxy for shareholder rights implies that corporate governance and firm value are significantly related in positive way. It’s most important to see the correlation between leverage and governance quality. As presented in Table 2, leverage is negatively related to corporate governance and significant, indicating a substitute relation between these two corporate governance mechanisms. Detail descriptions about other variables are shown in Table 1 and Table 2.
4.2 Regression analysis
The objective of our tests is to draw inferences about the relation between firm value, leverage, and the strength of shareholder rights, while controlling for a number of other factors.
The task is complicated because it can be argued that firm value, leverage, and corporate governance are all jointly determined.
14
We address the potentially endogenous relation among these variables (and account for other factors that could affect each of them) by estimating a three-stage least squares regression model. Moreover, because the test we construct is not just a cross- sectional test but also a panel data test, the most appreciate model might be the GMM model.
Equation (1) controls for firm size, measured as the natural logarithm of sales. We include the ratio of capital expenditures to assets as a proxy for growth opportunities (GO); however, if managers routinely overinvest, this ratio will instead pick up inefficient investment choices. The direction of GO’s coefficient is uncertain. Profitability measured as the ratio of earnings before interest and taxes to sales is also suggested to positively affect firm value. We also include two dummy variables, S&P500 and Delaware, in our valuation equation. Firms included in S&P500 are expected to experience better operating performance, and have positive influences on firm value.
The governance equation uses shareholder rights as the dependent variable, Tobin’s Q and leverage as simultaneously determined variables, and controls. According to Zheka (2006), some firm-level factors, like firm size and growth opportunity, could be the determinants of corporate governance. They suggest that firm size is significantly related to the level of corporate governance quality. We control over these two variables, firm size and growth opportunity, and use product market competition as the instrument variables following Gillan, Hartzell, and Starks (2003).
The leverage equation uses the debt-to-assets ratio as the dependent variable, shareholder right as the simultaneously determined variable. We include six determinants of leverage: firm size, profitability (EBIT-to-sales), fixed-asset ratio (net property, plant, and equipment divided by total assets), growth opportunity (the ratio of capital expenditures to total assets), research and development expenditures (divided by sales), and non-debt tax shield (depreciation and amortization divided by total assets). Shareholder rights could be negatively related to leverage, since managers of firms with bad corporate governance might use financial leverage to augment the assets under their control.
Table 3, 4, 5 reports the results of the OLS, 3SLS, and GMM estimation for the regression analysis from period 1998 to 2003. Table 3 shows the results of estimating for equation (1), while G-index and E-index are both used to proxy for corporate governance. For convenience to explain the quality of corporate governance, we use CG = 24-G-index and CG’ = 6-E-index to measure the strength of shareholder rights. As shown by Table 3, the coefficient of CG is
15
significantly positive in all estimation models, which indicates that firms with good corporate governance show distinctly higher firm value than firms with poor corporate governance. This result is consistent with GIM (2003), BCF (2005), and Cremers and Nair (2005).
In the best known test of an optimal capital structure model, Miller-Modigliani (1958) reported evidence of a positive relationship between firm value and leverage which they attributed to a debt tax shield effect. Our result is the same with the view of optimal capital structure model constructed by Miller-Modigliani. Table 3 shows a significantly positive coefficient on leverage both in 3SLS and GMM estimation model. Since we have known about the existence of estimation bias by OLS model in our investigation, the contrary direction of the coefficient on leverage is not surprising. The result that firm value and leverage are positively related follows from the fact that these two endogenous variables move in the same direction with changes in the exogenous factors (Hirshleifer and Thakor (1989), Harris and Raviv (1990a), Stulz (1990)). Therefore, leverage increasing (decreasing) changes in capital structure caused by a change in one of these exogenous factors will be accompanied by firm value increases (decreases).
The way how the scale of a firm relates to its firm value is uncertain on directions. As can be seen in Table 3, the coefficient of Size in GMM estimation model is significantly positive with the governance proxy of CG, consistent with the finding of Kadyrzhanova (2007). As mentioned before, growth opportunity is always positively related to firm value. But overinvestment by managers might affect firm value in undesired way, as GMM estimation model shows, using CG as a proxy for corporate governance. The same as we predicted before, profitability and inclusion in S&P500 have significantly positive influences on firm value.
Table 4 shows the analysis results of governance equation (equation (2)). Consistent with our finding on the relationship between corporate governance and firm value in previous analysis, firm value is significantly related to shareholder rights in the positive way. More important for our analysis, this significantly positive result is consistent under our three estimation model (OLS, 3SLS, and GMM) and robustness test of different governance definitions, which indicates that firm value and corporate governance do have strong influences on each other, supporting our endogeneity assumptions between central variables. As can be seen in Table 4, leverage is negatively related to corporate governance. Although the coefficient is not significant in 3SLS and GMM estimation model with CG’ as the proxy for corporate governance, we could have a general view about the relationship between corporate governance and leverage.
16
The coefficient of firm size is negative and highly significant. It implies that smaller size of firm is facing fewer anti-takeover provisions and associated with better corporate governance. As can be seen, growth opportunity is positively related to corporate governance. Masulis, Wang, and Xie (2006) argue that managers of firms operating in more competitive industries are less likely to shirk or put valuable corporate resources into inefficient uses. In other words, firms in competitive environment can contribute to improve governance quality. Consistent with our prediction, product market competition can affect corporate governance in positive directions, when CG’ is used to proxy for governance. Since governance index constructed by BCF (2005) is argued to be a more efficient proxy for governance than G-index, we should pay more attention on the results of analysis with CG’ when estimating equation (2).
Table 5 shows the most important results of our analysis on equation (3). The central idea of this paper is to determinant the way how corporate governance affects leverage. As showed in Table 5, the coefficient of governance is negative and highly significant in all estimation models, indicating that firms with higher governance quality exhibit lower levels of debt. Hence, the GMM regression results strongly support our hypothesized inverse association between leverage and the strength of shareholder rights. The result is consistent with the finding of Jiraporn and Gleason (2005). Our finding also supports the agency theory, which predicts that firms where shareholder rights are more limited (and, therefore, where agency costs are more acute) should adopt higher debt to mitigate the higher agency costs.
The leverage equation indicates that leverage is positively related to firm size. This finding is consistent with our predictions and with the findings of Kim and Sorensen (1986) for U.S.
firms. Warner and Ang, Chua, and McConnell provide evidence that suggests that direct bankruptcy costs appear to constitute a larger proportion of a firm's value as that value decreases.
It is also the case that relatively large firms tend to be more diversified and less prone to bankruptcy. These arguments suggest that large firms should be more highly leveraged. As Titman and Wessels (1988) point out, firms with assets that can be used as collateral may be expected to issue more debt to take advantage of this opportunity. Fixed-asset ratio has a positive but not significant influence on leverage, which is consistent with Marsh (1982), and Long and Hasbrouck (1988). Profitability, growth opportunity, R&D expenditures, and non-debt tax shields, as shown in Table 5, have significantly positive influences on debt ratio. The results are consistent with Bradley, et al. (1984), Castanias (1983), Long and Malitz (1985), Kester (1986), Marsh (1982), and Titman and Wessels (1988). These studies generally agree that leverage increases with fixed assets, non-debt tax shields, growth opportunities, and firm size
17