Optimal CEO Compensation with Search: Theory and Empirical Evidence




    Search for more papers by this author
    • Melanie Cao is with the Schulich School of Business, York University, and Rong Wang is with the Lee Kong Chian School of Business, Singapore Management University. An earlier version of the paper was circulated under the title “Search for Optimal CEO Compensation: Theory and Empirical Evidence.” This paper has been presented at Carnegie Mellon University, Central University of Finance and Economics, Fudan University, Queen's University, Shanghai University of Finance and Economics, the University of Toronto, the 2006 Southern Ontario Finance Symposium, the 2007 Northern Finance Association annual meeting, the 2008 Financial Intermediation Research Society annual meeting, the 2008 North American Econometric Society Summer Meeting, the 2008 Chinese International Finance Conference, the 2008 Financial Management Association meeting, the third annual conference on Asia-Pacific Financial Market, the 2009 European Financial Management Association meeting, and the 2011 meeting of Society for the Advancement of Economic Theory. We wish to thank the Editor, Campbell R. Harvey, an Associate Editor, and two anonymous referees for their thoughtful comments that have improved the paper significantly. We also thank Sugato Bhattacharyya, Douglas Blackburn, Neil Brisley, Douglas Cumming, Alex Edmans, Yaniv Grinstein, Brian Henderson, S. H. Seog, Shouyong Shi, Yisong Tian, Jan Zabojnik, and the seminar and conference participants for valuable comments and suggestions. Melanie Cao gratefully acknowledges financial support from the Social Sciences and Humanities Research Council of Canada.


We integrate an agency problem into search theory to study executive compensation in a market equilibrium. A CEO can choose to stay or quit and search after privately observing an idiosyncratic shock to the firm. The market equilibrium endogenizes CEOs’ and firms’ outside options and captures contracting externalities. We show that the optimal pay-to-performance ratio is less than one even when the CEO is risk neutral. Moreover, the equilibrium pay-to-performance sensitivity depends positively on a firm's idiosyncratic risk and negatively on the systematic risk. Our empirical tests using executive compensation data confirm these results.