The Implications of Complexity in CEO Pay Packages

Tying chief executive officer (CEO) pay to performance goals aims to solve a classic principal-agent problem, helping to ensure that the CEO acts in the best interest of shareholders. But can it be too much of a good thing – can compensation contracts be too complex?  Our research says yes.

Compensation contracts have numerous features, including different forms of compensation (e.g., salary, bonus, stock, and stock options), a variety of performance metrics, multiple periods for measuring performance, and different performance benchmarks (whether measured in relation to a specific pre-set target or the performance of a peer company). Compensation committees use these features to fine-tune incentives, but we argue that the resulting contracts can become so complicated that they contain conflicting incentives or that CEOs can suffer from information overload caused by having to focus on too many metrics. Both consequences may result in lower firm performance.

We define a contract as more complex when it contains more contingencies (that is, multiple factors for the manager to consider) and increases the CEO’s cognitive load. In assessing complexity, we consider the four dimensions of contracts mentioned above – types of pay, performance metrics, measurement periods, and benchmarks. We aggregate these features into a single measure, and, in our definition, the more features a contract contains, the more complex it is.

Using a sample of approximately 1,700 firms operating in the U.S. from 2006–2019, we find that contract complexity has, in fact, increased. This increase is driven by more performance metrics and more measurement periods as well as the greater use of benchmarks defined in relation to peer companies. We show that the components of pay more heavily weighted in CEO contracts (namely, short-term bonuses and stock awards) are more complex, suggesting that CEOs are exposed, in a material way, to complexity. That is, compensation contracts don’t just appear complex, they are economically complex.

We find that this complexity stems from factors related to the complexity of a company’s organization and operations and of the complexity of their peer firms’ compensation. Contract theory says that a simpler contract that allows for renegotiation may be preferable to an incomplete one in a complex environment. Although renegotiation is common in debt contracting, it’s uncommon in incentive contracts, as the possibility of renegotiation weakens incentives. We find evidence that contracts are more complex in firms that allow less discretion in their contracts, consistent with greater complexity being accompanied by a lower possiblity of renegotiation. Finally, we find that outside forces influence contract complexity, namely compensation consultants and the recommendations of proxy advisers like Institutional Shareholder Services. In the case of ISS, we find that, when words associated with contract complexity appear more frequently in ISS guidelines, firms’ contracts are more complex. Though not causal, this could be interpreted as the firm’s trying to secure ISS approval.

Why does complexity matter?  Investors, the media, and academics have all voiced concerns that CEO pay contracts have become so dense and detailed that executives struggle with them. Our findings suggest these concerns are valid. Firms whose CEOs have more complex contracts have lower future firm performance (measured by both accounting and stock returns). Our work illustrates the unintended consequences of compensation contracts with too many performance metrics or other provisions.

To understand why this happens, we consider how greater cognitive load imposed on the CEO by contract complexity might affect firm performance. We explore situations in which firms are exposed to extreme changes in industry growth – a condition under which the firm faces both challenges and opportunities requiring quick responses. We find that, when CEOs have more complex compensation packages, their firms perform worse under these conditions. The results are consistent with complex contracts diffusing managers’ attention and imposing cognitive costs on CEOs, thus hampering quick decisions. We also explore whether the negative association between complexity and firm performance is reduced when performance measures in the contracts are more correlated with each other. CEOs face lower cognitive load when performance metrics are highly congruent, as the actions a CEO takes to improve on one performance measure will end up also improving other measures. Indeed, we find this to be true; when the correlation between the performance metrics is high, the negative association between complexity and firm performance is weakened.

One might argue that contract complexity is a means of hiding higher CEO pay. If this were true, we would expect more complex contracts to have higher ratios of actual payouts from bonus contracts relative to target (expected) payouts. Our prediction assumes that firms disclose lower targets to suggest that the CEO is receiving reasonable pay, with the contract complexity obfuscating how higher payouts are achieved. But we find no association between actual-to-target payouts and contract complexity.  We would also expect that firms might use more sophisticated language in the compensation disclosure and analysis (CD&A) section of the proxy statement if hiding higher pay were their aim. However, we find no evidence of this either.

Research has extensively examined the determinants and effects of the individual features of CEO compensation contracts that independently contribute to complicated overall contracts. While the writing of contracts with several of these features might be well-intentioned, the outcomes generally are negative. Complex contracts are associated with lower future performance; the contracts appear to impede CEOs’ abilities to respond quickly to changing markets. The unintended consequences of contract complexity that we document, affirming concerns raised by investors and the media, suggest that boards should think twice before adding ever more features to CEO pay packages.

This post comes to us from Ana Albuquerque, an associate professor of accounting at Boston University’s Questrom School of Business; Mary Ellen Carter, an associate professor of accounting and EY Faculty Fellow at Boston College’s Carroll School of Management; Zhe (Michael) Guo, a PhD student at Boston University; and Luann Lynch, the Almand R. Coleman Professor of Business Administration at the University of Virginia’s Darden School of Business. It is based on their recent paper, “Complexity of CEO Compensation Packages,” available here.