|
|
||
|
|
||
|
|
|
||
|
|
||
|
|
The Agency Problems of Hedging and Earnings Management
Lin Nan
Using a principal–agent model with a mean-preserving spread hedging structure, this paper studies the interaction between hedging and earnings management decisions. The key assumption is that hedging makes earnings management more costly. In this model, hedging is efficient if the hedging decision is contractible, because it reduces both the risk premium and the equilibrium amount of earnings management. However, when the hedging decision is not contractible, a strategy of discouraging hedging but allowing earnings management may be optimal, because encouraging hedging may require a more costly compensation scheme to compensate the agent for reduced earnings management. Here, encouraging truth-telling is costly and inefficient when there is no opportunity to hedge. However, with hedging, the cost of motivating truth-telling can be reduced. In some situations, a strategy of motivating hedging and truth-telling is optimal. Hedging and earnings management are strategic substitutes in this paper, which is consistent with existing empirical evidence. This study also provides an alternative explanation for that evidence.
In this paper, a risk-neutral principal contracts with a risk-averse manager for two periods. The output in each period, which is the manager's performance measure, is a function of the manager's effort, the profitability parameter of that period, and a noise term. The manager chooses his effort level at the beginning of each period. At the end of the first period, the manager privately observes the first-period output and decides whether to manipulate the reported first-period output at a personal cost. The personal cost is increasing in the amount of manipulation and decreasing in the noise term. He also has an option to hedge in the second period. The key assumption is that hedging reduces the noise in the second period, thus improving the informativeness of the performance measure, but it also makes earnings management more costly. The principal compensates the manager on the basis of the manager's reported outputs at the end of the second period.
The paper shows that when the decision to hedge is contractible, hedging alleviates the agency problem by reducing both the risk premium and the equilibrium amount of earnings management. However, if the decision is not contractible, hedging will not always ease the agency problem. The reason is that hedging increases the manager's personal cost of earnings management and, consequently, the principal's cost to compensate the manager. Motivating hedging may require a more consistent compensation scheme across periods to attenuate the appeal of earnings management, which may be inefficient. In some situations, a strategy of discouraging hedging but allowing earnings management is efficient. In addition, this paper shows that the principal will tolerate some earnings management when there is no hedging option, because it is costly to motivate truth-telling. However, with the hedging option, a strategy of motivating truth-telling and hedging may be optimal, because hedging can significantly lower the cost of eliminating the deadweight loss from earnings management. In this paper, the technology and endogenous incentives lead to a relationship of strategic substitutes between hedging and earnings management.
Although there have been no theoretical studies on the relationship between hedging and earnings management, the subject has been explored in a few recent empirical studies. Barton (2001) reports a significant negative association between the use of derivatives and the magnitude of discretionary accruals, which serves as a proxy for earnings management. Pincus and Rajgopal (2002), focusing on the oil and gas industry, report similar results. Both papers speculate that hedging and earnings management are alternative mechanisms for smoothing earnings. Consistent with the empirical evidence, hedging and earnings management are strategic substitutes in this paper. However, unlike the explanation proposed by Barton 2001 and Pincus and Rajgopal 2002, this paper's substitutional relationship is achieved through hedging's effect on the cost of earnings management and endogenous incentives, which provides an alternative (or complementary) explanation for that evidence.