A Learning-Based Approach to Evaluating Boards of Directors

Abstract

Uncertainty about a firm’s governance is reflected in its stock return volatility. A model of learning about director ability implies that the magnitude of the decline in volatility over a director’s tenure is a function of his value impact. This paper uses this prediction from the theory to develop a learning-based approach to evaluating corporate boards. Its estimates suggest that directors have real value effects and that governance accounts for a substantial part of stock return volatility. I use this novel framework to study the marginal value of different kinds of directors, thereby revisiting the literature on boards and testing new hypotheses. Among other findings, I find that the marginal value of independent directors depends on the degree to which firms are insulated from the market for corporate control and that incoming directors on well compensated boards have more impact on firm value.