Malmendier, Ulrike, and Geoffrey Tate. "Superstar CEOs." Working Paper, Stanford University. May 9, 2005.

This strong study reviews the performance of CEOs "who achieve 'superstar' status via prestigious nationwide awards from the business press," and finds that they "subsequently underperform beyond mere mean reversion, both relative to the overall market and relative to a sample of 'hypothetical award winners' with matching firm and CEO characteristics."

Superstar CEOs are first identified using hand-collected data for the 1975-2002 time period. Indications of star status are derived primarily from magazine awards, notably from Business Week and Financial World. Award-winning CEOs are then matched with non- Award-winners on the basis of firm characteristics and performance (data sources include: Compustat Execucomp database, Compustat, CRSP, and the Fama-French portfolios). The authors also measure "the CEO's propensity to undertake tasks that distract from maximizing profits," such as writing books.

The dependent variables are stock returns and return on assets (ROA):

Abnormal stock returns are measured over the 11 days, 1 year, 2 years, and 3 years following the announcement of the awards. The authors find no short-term effect, but discover significant stock underperformance over all three long-term time periods: "firm performance, measured using stock returns data, is lower once a CEO attains celebrity status."

The authors also discover "a pronounced decline in return on assets" once CEOs achieve celebrity status... Over the three years following an award year, ROA is roughly two and a quarter percentage points lower than in the year preceding and year of the CEO award."

Extensive analysis suggests that these financial results are not merely the product of mean reversion.

The study also finds important impacts on executive compensation. Although award winners do not earn notably greater cash compensation, they receive significantly larger overall pay packages than CEOs with similar performance.

Finds that CEO performance is worse and compensation is higher in firms with weaker corporate governance.

This study was presented at the joint Stanford - Berkeley seminar on May 11th, 2005. A copy of the paper is available here: http://www.haas.berkeley.edu/finance/superstarceos_09may2005.pdf

See also McMillan (2005).