CEO Compensation: Agency Theory is Irrelevant but not the Neoclassical Game-Theoretic Framework

Often criticized in the civil society for its magnitude, though considered with mixed appreciations by academics, CEO pay has been objects of many contributions. Reviewing key papers that have raised controversies, we discuss divergent viewpoints with simple game theoretic models in the neoclassical spirit. We assert the complete inadequacy of the agency and asymmetry of information models for explaining CEO compensation, but we diverge from those who reject the optimal contracting approach and show how reasoning with the classical tools of utility maximization, rationality, freedom to participate, and price sets on markets, competitive or not, can model a broad range of situations, including those put forward as arguments against the microeconomic approaches of compensation. The CEO-Board relationship should not be studied as a delegation issue within a hierarchical organization with the shareholders sitting at the top, but rather as a market bargain in search of optimal contracting with symmetrical position and information of both parties, but with asymmetrical reservation utility because the distribution of talent to manage is itself highly asymmetrical, expressing the game power of each side.