Date of Award
Doctor of Business Administration (DBA)
Jensen and Meckling (1976) and Jensen (1986) argue that the separation of ownership and control may generate agency problems between managers and shareholders. The equity-based compensation, by tying managerial wealth to firm long-run stock performance, can incentivize managers to be more receptive to undertaking value-increasing financial policies and to improving firm performance therefore can be used as an effective tool to achieve consonance between managers actions and shareholders interest. Over the last two decades, the increased prevalence of equity-based compensation in the form of stock and options, is partially due to an increased acceptance of the alignment effect of equity-based compensation (Murphy, 1999; Perry and Zenner, 2000; Hall and Leibman, 1998; Jensen and Murphy, 1990).
The current literature examining the effect of CEO pay-performance sensitivity on corporate payout policy and firm value produces inconclusive findings. In this dissertation, I revisit the effect of CEO pay-performance sensitivity on payout policy and firm performance. My dissertation consists of two essays. The first essay examines the effect of CEO pay-performance sensitivity on payout policy, including dividends and share repurchases, for a sample of 6,656 firm-year observations from 1995 to 2008. The second essay examines the relation between CEO pay-performance sensitivity and firm value for firms with different value for a sample of 19,313 firm-year observations from 1992 to 2008.
To increase the power of tests of the effect of equity-based incentives on payout policy and firm value, I use the dollar change in CEO wealth for a percentage change in firm value scaled by CEO's total annual pay as a proxy for CEO pay-performance sensitivity (Edmans et al., 2009). Percent-percent incentives equal dollar-dollar incentives (fractional CEO ownership) multiplied by firm value then scaled by CEO wage. This measure indicates that a small effective ownership can induce high level of incentives since firm value is substantially greater than the CEO's wage. Furthermore, in the first essay, I employ a Fama and MacBeth (1973) procedure to reduce bias resulting from heteroskedasticity and autocorrelation in the regression residuals. In the second essay, I employ quantile regression methodology. Quantile regression models the relation between CEO pay-performance sensitivity and specific percentiles of the firm value, providing an opportunity to examine how changes in CEO pay-performance sensitivity affect the entire distribution of the firm value.
In the first essay, I document that the likelihood of dividend payout is decreasing in CEO pay-performance sensitivity for the full sample and the subsample of firms that did not pay dividends in the previous year. This evidence is consistent with a substitution effect between dividend payout and CEO stock compensation. Conversely, I find that the propensity to pay dividends is increasing in CEO pay-performance sensitivity for the subset of firms that paid dividends in the previous year. For firms with positive dividends payout, the weight on the need to reduce the potential increase in external financing is greater than the weight on the need to reduce agency problems. I also find that the level of dividend payout is decreasing in CEO pay-performance sensitivity for dividend paying firms. The negative relation is nonlinear, stronger for CEOs with higher pay-performance sensitivity, and stronger for firms with greater potential for free cash flows. Lastly, I document that dividend smoothing decreases in CEO pay-performance sensitivity. CEOs with greater equity-based incentives tend to deviate from past dividend levels to accommodate future investment opportunities. My results are robust to alternative sample selection procedures and model specifications that address causality. My models also perform well in both in-sample and out-of-sample predictions.
In the second essay, I document a significantly inverted-U relation between CEO pay-performance sensitivity and Tobin's Q for all percentiles of the conditional distribution of Tobin's Q. Moreover, the strength of the coefficients on CEO pay-performance sensitivity and its square are significantly increasing in the percentiles of the conditional distribution of Tobin's Q. The difference in the coefficient estimates is also statistically significant across the entire distribution of Tobin's Q. This suggests that the incentive alignment effect of higher CEO pay-performance sensitivity is more pronounced in firms with higher firm value, but so is the risk aversion effect of excess managerial ownership. My results are robust to alternative methodologies to control for unobservable firm fixed-effects, alternative model specifications that include proxies for board quality and effective monitoring, and alternative specifications that address causality.
Liang, Hui, "" (2011). Dissertation. 386.