Compensation Goals and Firm Performance

In the article Compensation Goals and Firm Performance which is forthcoming in the Journal of Financial Economics, we study the growing use of specific performance goals in top executive compensation packages. A recent survey by the consulting firm Hay Group found that more than half of the CEOs in their study have compensation tied to explicit goals, up from around 35% just four years earlier. Prominent shareholders like Warren Buffet agree with the need for such targets, stating: “Lacking such [goals], managements are tempted to shoot the arrows of performance and then paint the bull’s-eye around wherever it lands.”

A typical equity or non-equity grant linked to firm performance identifies threshold, target and maximum values for one or more accounting or stock price-based metrics. The payout from the grant or the vesting schedule of the grant is then tied to the firm achieving these particular performance goals. For example, a manager may receive no payout if performance is below the threshold and her payout may increase as performance exceeds the threshold. The slope of the pay-performance relationship (PPR) may also change at the target and the maximum value, with discontinuous slope changes generating a “kink” in the PPR. We use a comprehensive dataset containing information on the performance goals employed in pay contracts obtained from IncentiveLab to highlight some of the costs of this popular pay feature.

Rewarding managers for achieving explicit performance goals certainly has a bright side. It makes pay more transparent and offers strong incentives, especially when the goal is challenging. Goals may also force the board to think about the firm’s strategic imperatives and direct managerial attention towards those. On the other hand, identifying explicit performance targets and having “jumps and kinks” in the PPR at the target may also have a dark side. Managers may try to beat the target either to obtain the payout (when there are discontinuities in the PPR) or to avoid punishment by the board for perceived underperformance. In an effort to influence subsequent targets down, managers may also wish to avoid beating the target by much, the “target ratcheting” effect.

We begin our empirical analysis by comparing the firm’s actual performance to the target performance in CEO compensation contracts. We find that firm performance is more likely to beat the target by a small margin as compared to fall short by a small margin. The discontinuity in the density of firm performance is consistent with firms managing reported performance to just beat the target. We find that performance clustering at the target is greater if the grant is contingent on a single metric as opposed to multiple metrics. This is consistent with managers not being able to “just beat” multiple goals that may be correlated with each other. We also document greater performance clustering if the rewards for exceeding the target are not high: i.e., in the case of grants involving a concave kink at the target and when the grant involves cash as opposed to stock payout.

We find no performance clustering for grants contingent on sales goals while there is significant clustering for EPS and profit goals. This is consistent with it being difficult to make up a shortfall in sales as opposed to a shortfall in profits. We also do not find any performance clustering when the grants are contingent on relative performance goals.

In our subsequent analysis we try to address two questions in the context of our findings: why does reported performance cluster at the target? and how do managers report performance at just above the target?

We find that current performance has a strong and positive association with subsequent targets. This provides one clue as to why performance may cluster at the target. Managers may be reluctant to report performance much above the target as that would imply higher future targets. We also document that CEOs are more likely to be fired if firm performance falls short of the target. Thus, boards appear to evaluate CEO performance relative to the target and fire underperforming CEOs. This may provide another reason for the CEO to ensure firm performance exceeds the target.

Depending on the metric involved, firms can employ a variety of means to meet a goal. They can both manage reported income and also discretionary expenditures to meet the goal. We find that firms do a bit of both. Firms with reported performance just above EPS goals have much higher accounting accruals as compared to firms with reported performance just below EPS goals. We also find that firms with reported performance just above profit based goals have much lower R&D and SG&A expense as compared to firms with reported performance just below profit goals.

Our article highlights that the growing trend of specifying explicit performance goals in pay contracts has certain subtle costs that firms should take into account. Our study also points to ways in which boards can minimize such costs. As far as possible, boards should link managerial pay to multiple as opposed to single metrics as they provide reduced incentives to manage performance. Providing a continuous link between pay and performance and avoiding discontinuities and kinks in the relationship is another way firms can avoid distortions. Finally, evaluating managers on relative performance goals is an easy way for firms both to reduce the risk imposed on the manager and also to avoid distortions from absolute performance goals.

The complete article is available here.