Bargaining Dynamics and CEO Pay Under Fat-Tailed Firm Distributions
Abstract We study a dynamic contracting problem in which a principal designs an incentive contract when the agent’s bargaining weight is a state variable that evolves with realized performance. We extend a recursive bargaining-drift framework by allowing firm output to follow a bounded Pareto distribution, which generates tractable fat-tailed risk while preserving the monotone likelihood ratio property and related regularity conditions used to implement incentive compatibility. The optimal contract is characterized numerically via value function iteration on the bargaining state, and we quantify how tail thickness and drift intensity reshape the policy rules for compensation, continuation values, and the implied degree of insurance. Fat-tailed output increases state dependence and amplifies persistence in bargaining power, producing stronger feedback from extreme outcomes into future contract terms. In the data, we estimate a reduced-form analogue of the bargaining-drift law of motion using quantile regressions on U.S. executive compensation and show that the estimated responsiveness of incentive-heavy pay varies systematically across the compensation distribution in ways consistent with the model’s state-dependent predictions.