Performance Terms in CEO Compensation Contracts

CEO compensation in U.S. public firms has attracted a great deal of empirical work. Yet our understanding of the contractual terms that govern CEO compensation and especially how the compensation committee ties CEO compensation to performance is still incomplete. The main reason is that CEO compensation contracts are, in general, not observable. For the most part, firms disclose only the realized amounts that their CEOs receive at the end of any given year. The terms by which the board determines these amounts are not fully disclosed.

The fact that the contractual terms are not fully observable has led researchers to doubt that such contracts optimally tie CEO compensation to performance. For example, Bebchuk and Fried (2003) argue that companies have decoupled compensation from performance and camouflaged both the amount and performance-insensitivity of pay. Morse et al. (2011) show both theoretically and empirically that, with lack of transparency of compensation contracts, powerful managers have the ability to rig their performance-pay for their own benefit.

In December 2006, the Securities and Exchange Commission (SEC) issued new disclosure requirements on CEO compensation. These requirements came as a response to investor concerns that in recent years CEO compensation packages have not been properly disclosed or well understood. According to these new requirements, firms now must provide additional information about the contractual terms of their compensation to the CEO. In particular, firms need to disclose the types of performance measures that they use to determine CEO rewards, the performance targets, and the performance horizon.

In our paper, Performance Terms in CEO Compensation Contracts, forthcoming in the Review of Finance, we use this newly available data to examine how firms tie CEO compensation to performance and the extent to which such practices support the predictions of optimal contracting theories. We focus on performance-based awards, since these awards are the ones where full disclosure of the rationale behind the award is available. Nevertheless, we also consider other types of awards in our robustness analysis. We first document the choice across the wide array of performance measures and then we examine the relation between these performance measures and firm characteristics. Our sample consists of firms in the Standard and Poor’s (S&P) 500 index in fiscal 2007. We collect information from the proxy statements on the performance measures used in the performance-based awards in fiscal year 2007. We focus on identifying the different types of performance measures and their relative weights. We observe that 90% of our sample firms grant some type of performance-based awards. The average value of these awards is 4.8 million dollars.

In general, firms pre-specify their performance goals over several performance measures. On average, 79% of the estimated value of performance-based awards is based on accounting-performance measures, 13% is based on stock-performance measures (i.e., market-based), and 8% is based on non-financial measures. Firms use a wide array of accounting measures. Firms reward CEOs based on income measures (e.g., earnings-per- share (EPS), net income growth, and earnings before interest and taxes (EBIT)), sales, accounting returns (e.g., return on equity, return on assets), cash flows, margins, cost-reduction measures, and Economic Value Added (EVA)-type measures. On average, 56% of the estimated value of performance-based awards assigned to accounting measures is tied to income measures. A significant portion of the awards is also assigned to sales measures (12%) and accounting returns measures (17%). We find that larger firms and firms with larger growth opportunities tend to rely more heavily on market-based measures, and firms that are more mature tend to rely more heavily on accounting-based measures. In addition, among accounting measures, sales are used by firms with larger growth opportunities and accounting returns are used more heavily by more mature firms with fewer growth opportunities. We also find that firms in similar sectors tend to adopt similar performance measures.

Overall, our findings regarding the relation between firm characteristics and performance measures suggest that firms tend to choose performance measures that are more informative of CEO actions. In growth firms, where CEO optimal actions are improving long-term growth opportunities, end-of-year accounting performance measures are likely to be less informative of optimal CEO actions. For these firms, stock price performance, which captures investors’ perception regarding firms’ long-term growth opportunities, is a more informative measure. Among accounting measures, growth firms tend to rely on sales growth measures, which again capture CEO actions associated with growth. In contrast, in mature firms, where CEO focus is on maximizing value from existing operations, end-of-year accounting performance measures are more informative of CEO actions. Among accounting measures, firms tend to rely on accounting returns, capturing efficiency in allocation of capital to existing operations. Our evidence is largely consistent with the informativeness principle, which emerges in optimal contracting theories such as Holmstrom (1979).

Our study contributes to the existing literature in several ways. First, the disclosure rule allows us to document the large array of performance measures that are used in CEO compensation contracts and to examine firms’ choices across the different measures. With the new data, we are able to directly examine the choice of different performance measures in CEO compensation contracts and relate it to contracting theory. Past studies could not observe the choice of performance measures across the different components of compensation contracts because this data was not available. As a result, most studies have estimated the choice of performance measures from observed compensation outcomes. Few previous empirical studies had access to more precise data regarding the terms of the contracts, but even then, the data was available only for particular components of the contract. Second, the rich information on the variety of performance measures allows us to shed new light on the relation between contractual choices and firm characteristics and to tie our findings to existing theories.

Our focus in this study is on awards whose performance terms are pre-specified. But we note that not all types of CEO awards are pre-specified. For example, firms can give discretionary end-of-year bonuses and they can decide to award CEOs with options or stocks at their discretion. Unfortunately, we cannot identify the reasons behind these awards and therefore we exclude them from the main analysis. This is a limitation of our study. However, we do examine whether these discretionary awards are complements or substitutes to the pre-specified performance-based award. We do not find any significant relation between discretionary awards and pre-specified performance awards. This means that this portion of compensation is given for reasons other than performance, (e.g., retention purposes).

Our study produces two interesting findings that require further examination. First, a large portion of CEO awards is given at the discretion of the board. How exactly this portion of the awards is determined is an interesting topic for future research. Second, we find that CEO shareholdings have little association with the level of market-based awards in the CEO contract. This result is puzzling because we expect CEO shareholdings to act as a substitute to the market based awards. We believe that further investigation of this result is another fruitful area for future research.

The full paper is available for download here.