Corporate Hedging and the Design of Incentive-Compensation Contracts

Sterling Huang is Assistant Professor of Accounting at Singapore Management University. This post is based on a recent paper authored by Professor Huang; Chris Armstrong, EY Associate Professor of Accounting at The Wharton School of the University of Pennsylvania; and Stephen Glaeser. Related research from the Program on Corporate Governance includes: Paying for Long-Term Performance by Lucian Bebchuk and Jesse Fried (discussed on the Forum here).

The theoretical agency literature highlights the importance of risk-related agency conflicts—whereby undiversified executives are more averse to firm-specific risk than are diversified shareholders—as a potential source of wealth destruction. Although providing executives with incentives tied to stock price can sometimes alleviate these agency conflicts, doing so exposes them to risk that requires payment of a commensurate risk premium. Consequently, firms trade off the benefits of providing incentives against the costs of compensating executives for bearing the associated risk. While this tradeoff leads to relatively straightforward predictions about the effect of risk on executives’ compensation, the effect of risk on executives’ incentives is theoretically ambiguous (Jenter, 2002; Prendergast, 2002; Hemmer, 2006, 2012).

In our paper, Corporate Hedging and the Design of Incentive-Compensation Contracts, we examine how executives’ ability to hedge risk that was previously difficult and costly to manage influences the design of their incentive-compensation contracts. Identifying the effect of risk on the design of executives’ incentive-compensation contracts is empirically challenging not only because of these conflicting theoretical predictions, but also because of the endogenous relation between risk and incentives. To overcome this concern, we examine a natural experiment involving the introduction of exchange-traded weather derivatives on firms in the utility industry.

Prior to the introduction of weather derivatives, it fiwas difficult (i.e., costly, if at all feasible) for these firms to hedge the risk associated with their exposure to weather fluctuations. Weather derivatives were a financial innovation that allowed firms to hedge these risks for the first time or, at a minimum, significantly reduced the cost of hedging weather-related risk. Consequently, the introduction of weather derivatives allowed firms to alter their exposure to weather risk. We find that firms with greater historical exposure to weather risk are more likely to use weather derivatives to hedge their exposure to this risk. In particular, these firms experience a statistically significant and economically meaningful reduction in the covariance between their stock returns and weather-related outcomes following the introduction of weather derivatives. We corroborate this indirect evidence of increased hedging by searching our sample firms’ 10-K filings for references to weather derivative contracts and find that these firms are also significantly more likely to discuss their use of these contracts. Together, these two findings suggest that our sample firms used weather derivatives to hedge at least some of their exposure to weather risk and experienced a meaningful reduction in their exposure to weather risk as a result.

These changes in firms’ exposure to weather risk should have influenced the design of their executives’ incentive compensation contracts in several important ways. First, the ability to hedge risk should affect the amount of executives’ annual compensation. A portion of an executives’ annual pay consists of a risk-premium to compensate them for bearing the risk associated with their performance-based incentives and firm-specific human capital. If hedging eliminates some of this risk, then executives will be exposed to less firm-specific risk and, consequently, should demand and receive less of a risk premium in their annual pay. Consistent with this prediction, we find that the introduction of weather derivatives caused a significant reduction in the total annual compensation—including both the cash and equity grant components—of the CEOs of firms with greater historical exposure to weather risk. We attribute this reduction in annual compensation to a decrease in the risk premium that these CEOs receive for having their wealth (e.g., equity holdings and human capital) exposed to weather risk.

Second, the ability to hedge risk should also affect executives’ incentives in general and their equity incentives in particular. Although the prevailing view in the empirical literature is that risk should have a negative relation with the strength of executives’ incentives, numerous theoretical studies make it clear that the relation is theoretically ambiguous (Holmstrom, 1979, 1982; Jenter, 2002; Hemmer, 2006, 2012; Guo and Ou-Yang, 2015). Although this point has been made in several theoretical (e.g., Jenter, 2002; Hemmer, 2006, 2012) and empirical studies (e.g., Demsetz and Lehn, 1985; Core and Guay, 1999, 2002), the belief that risk should have a negative relation with incentives is still common.

The intuition for the theoretical ambiguity of the relation between risk and incentives is simple: when the moments of a performance measure (e.g., stock price) are correlated, the executive’s actions cannot affect its mean without also affecting its variance. However, the highly stylized “standard model” that is frequently cited for the predicting a negative relation between risk and incentives fails to capture this endogenous relation.

We find that the introduction of weather derivatives caused a significant decline in the equity incentives of the CEOs of firms with greater historical exposure to weather risk. This finding, coupled with our evidence that firms with greater historical exposure to weather risk experienced a significant reduction in their risk, is evidence of a positive, rather than a negative relation between risk and incentives. Although this finding is entirely consistent with theoretical predictions from some of the most well-known agency models (e.g., Holmstron, 1979; Holmstrom and Milgrom, 1987), it is inconsistent with empirical studies that predict a negative relation between risk and incentives (e.g., Aggarwal and Samwick, 1999).

The full paper is available for download here.

Both comments and trackbacks are currently closed.