CEO Compensation Contract Design under Moral Hazard and Adverse Selection

Abstract

In modern corporate governance, the separation of ownership and management rights leads to increasingly prominent principal-agent problems. Information asymmetry leads to adverse selection in advance and moral hazard after the event, which significantly increases agency costs and weakens the incentive effect of enterprises. In view of the realistic dilemma of unobservable executive ability and unverifiable efforts, the traditional single performance-linked compensation model is difficult to adapt to the differentiated needs of different types of executives. Therefore, this study focuses on the design of CEO compensation contract, and constructs a theoretical model that includes both moral hazard and adverse selection. It aims to achieve type separation and incentive compatibility through mechanism design, so as to maximize shareholders ' interests under the condition of information asymmetry. Based on the principal-agent theory framework, this study sets shareholders as risk-neutral and CEOs as risk-averse agents, and constructs a single-period model including hidden ability types and hidden effort levels. By introducing a linear compensation contract and a mean square utility function, the CEO 's optimal effort decision is internalized, and a contract menu is constructed based on participation constraints and incentive compatibility constraints. Aiming at the problem that the model is difficult to solve analytically, a two-stage grid search method is used for numerical optimization. Firstly, the high-yield area is located by global scanning of coarse grid, and then the optimal incentive coefficient and fixed compensation combination are determined by local fine search of fine grid. This method effectively improves the efficiency and stability of the solution, and has good expansibility. The results show that the optimal compensation contract shows significant differentiation characteristics : high-capacity CEOs correspond to high performance sensitivity coefficient (b _ H * = 0.837) and negative fixed compensation (w _ H * = -0.638), forming a ' strong incentive and high deposit ' model ; the CEO with low ability corresponds to a lower incentive intensity (b _ L * = 0.327) and a fixed salary close to zero. This design effectively compresses the information rent of high-ability CEOs, and at the same time promotes the two types of CEOs to truthfully reveal their own types and make corresponding efforts to achieve incentive compatibility and type self-selection. This study not only expands the contract theory under dual information asymmetry, but also provides theoretical basis and practical reference for enterprises to design differentiated executive compensation and regulators to formulate reasonable incentive policies. In the future, it can be further extended to continuous type setting, multi-task performance indicators and dynamic contract scenarios to enhance the realistic explanatory power of the model.