Print
E-MailThis post comes to us from Murray Carlson, Associate Professor of Finance at the University of British Columbia, and Ali Lazrak, Associate Professor of Finance at the University of British Columbia.
CEO compensation typically includes both performance sensitive and performance insensitive components. This pay structure can be readily rationalized in a contracting setting (e.g., Holmstrom (1982)). Performance sensitive payments link a manager’s value enhancing actions to his wealth thereby aligning his incentives with those of the firm. These payments are risky, however, and in contracts with a risk averse manager risk sharing motives give rise to a role for the performance insensitive component. Compensation structure has a direct impact on manager objectives and it is natural to expect it to affect a manager’s choice of firm risk and, consequently, the dynamics of the firm’s security prices. Aspects of these linkages have been explored empirically, but the structural modeling of the impact of CEO pay on risk choice, capital structure, and the pricing of financial securities remains relatively unexplored.
In our paper, Leverage Choice and Credit Spread when Managers Risk Shift, forthcoming in the Journal of Finance, we demonstrate the relevance of the agency costs of Jensen and Meckling (1976) for structural models of leverage choice and credit spreads. Assuming a realistic compensation structure for risk-averse managers, consisting of cash and stock, we show that managers will optimally choose to lever the firm and that their resulting pay will be convex in the firm’s terminal liquidating pre-tax payout. This convexity induces asset substitution, leading to riskier payouts and higher credit spreads than predicted by the prior literature. We also demonstrate that optimal leverage choice is the result of a balance between tax benefits and the utility cost of ex-post risk shifting. Our work thus highlights that operating behavior induced by compensation terms can be of first-order importance for understanding debt levels and credit spreads.
To empirically evaluate our model, we use a large cross-section of 608 US based corporate credit default swaps covering 2001 to 2006. We confirm the model’s prediction of a significantly positive relationship between cash-to-stock ratios and CDS rates when CEO salary is used as a proxy for cash pay and CEO stock holdings are used as proxies for stock pay. Reasonable variation in proxies for cash-to-stock result in variation in CDS rates of up to 21% even after controlling for the traditional structural determinants of spreads such as leverage and stock volatility. We further find that CEO cash-to-stock is informative for leverage choice. Our empirical results are not necessarily causal since cash-to-stock and capital structure can be jointly determined. The exploratory empirical associations that we document in this paper should, however, help guide further empirical research that more directly addresses potential endogeneity in the data.
Our empirical work contributes to the extensive literature on the determinants of credit spreads (e.g., Campbell and Taksler (2003), Collin-Dufresne, Goldstein, and Martin (2001), and Elton et al. (2001)). Our focus on CEO cash-to-stock is new in the credit spread literature, but relates to recent empirical studies on the impact of CEO pay structure, rather than CEO pay levels.
The full paper is available for download here.
