Three Essays on Executive Incentive Pay and Causal Inference in Corporate Finance

<p>Determinants of CEO Compensation and Firm Performance — A Survey</p> <p>Reviewing a large set of literature on CEO compensation and firm performance, I find the following. First, institutional and macro-level factors tend to take the first order importance in determining CEO compensation and firm outcome as well as many other factors in corpo- rate finance and corporate governance. Second, relations among the factors that determine CEO compensation and firm outcome are complex and often the exact causal directions are ambiguous. Third, endogeneity is present and it’s likely many factors including CEO compensation and firm performance are jointly determined. Fourth, CEO compensation has been mainly characterized in a principal-agent framework and many factors are regarded as either constraining or exacerbating the agency problems.</p> <p>Optimality of Dynamic CEO Compensation and the Endogeneity Problem</p> <p>Optimal dynamic contract theory postulates that the CEO’s portfolio needs to be rebal- anced each year to maintain the optimal incentive that implements a targeted effort. Thus,</p> <p>given previous year’s incentive, there exists an optimal adjustment that has to be made by the principal in order to maximize firm value. If contracts are optimally written and im- plemented, it’s expected that the derivative of firm value on changes in CEO incentive is zero. I test this hypothesis by estimating the treatment effect of changes in CEO incentive on firm value while controlling for a large set of lagged market, firm, board and CEO charac- teristics and using the inverse probability-of-treatment weighting method. I also use lagged firm policy variables as proxies for unobserved characteristics such as CEO risk aversion. Part of the results show that there is no significant effect of incentive adjustment on firm value, thus possibly confirming the optimality of CEO incentive contracts. However, there is also evidence that unobserved confoudning exists and the inherent endogeneity may bias the results.</p> <p>Causal Inference in Corporate Finance</p> <p>I advance a methodology for sound causal inference in corporate finance. I first review three applications of the quasi-experimental methods to study causal relations in corporate finance and argue that ensuring the validity of the quasi-experimental design ironically re- quires a good amount of knowledge about the underlying causal mechanisms which is so far lacking. Then I argue that the structural equations model provides a systematic framework for causal inference and the complexity and interconnected relations in corporate finance can be an advantage in learning causal relations by applying causal discovery techniques. I also present three methods to estimate causal effects once a better understanding of the under- lying causal structure in corporate finance is achieved. Lastly I apply one of the methods to study the marginal contribution of CEO incentive in explaining firm performance beyond a large set of market, firm, board and CEO characteristics as well as lagged firm perfor- mance. I find that change in CEO incentive, or the incentive adjustment, has persistent and strong predictive power over Tobin’s q, which lends some support to the dynamic contracting theory.</p>

Name thesis:degree_name

Philosophy, PhD

Level thesis:degree_level

Restricted to Claremont Colleges Dissertation

Discipline thesis:degree_discipline

School of Politics and Economics

Year dc:date.available

2020

Author dc:creator

• Wang, Xiaolong

Contributors dc:contributor

• C. Mónica Capra
• Darren Filson
• Ying-ying Lee

• CEO compensation
• Endogeneity
• Firm performance
• Incentive
• Inverse probability-of-treatment weighting
• Structural equations model

Repository record dc:identifier

https://scholarship.claremont.edu/cgu_etd/307

OAI identifier oai:identifier

oai:scholarship.claremont.edu:cgu_etd-1269