This paper presents estimates from a latent variables model of the relation between corporate governance and financial performance. We use data on large US corporations to estimate the correlation, conditional on the firms' investment opportunity set, between governance and performance. We find that this correlation is statistically speaking zero. This result is consistent with the equilibrium view (Demsetz, 1983) in which firms optimize corporate governance arrangements subject to the constraints imposed by the investment opportunity set, such that observed corporate governance arrangements and firm performance are uncorrelated. The intuition behind this statement is that, if governance and performance were correlated, performance could be improved by making changes to the governance arrangements, which is at odds with an equilibrium situation.
corporate governance firm behavior hypothesis testing errors in variables