The Stock Market Bubble, Shareholders' Attribution Bias and Excessive Top CEO Pay

摘要

I use aggregate time series data on profits in the corporate sector, Standard and Poor's Composite Stock Price Index, and the average total pay of the top 100 CEOs to look for evidence in favor or against a fundamental attribution bias-based explanation of the recent explosive growth in CEO pay. I hypothesize that shareholders overattribute prominent increases and decreases in the prices of corporate stocks to the leadership and skill of the CEOs and underplay the role of stock market fluctuations, which are beyond CEO control. I recast this hypothesis as a simple endogeneity test, and I cannot reject the view that market fluctuations mechanically cause changes in CEO compensation with no detectable reverse causality. Further I find that, in the aggregate data increases in CEO pay decrease corporate profits. To complement the time series evidence, I use 4-factor model risk-adjusted returns as a direct measure of CEO skill. I show that in the cross section the relationship between individual CEO pay and skill is very weak (economically small and statistically insignificant). I conclude that in the late1990s stock market bubble period shareholders were taken for a ride and ended up paying huge amounts of money to their CEOs for no rational reasons.