Essays in financial economics

Thesis

My thesis consists of three individual papers, covering topics on corporate social responsibility, public policies, corporate governance, and econometric methodology. The first paper explores the interaction between public policies and private-sector prosocial motives. Using the case of greenhouse gas emissions, I show that state-level emissions reduction targets adopted by nine U.S. states lower shareholder pressure on firms to cut emissions, as evidenced by fewer emission-related shareholder proposals and lower voting support rates. Furthermore, these state-level targets do not reduce corporate emissions as intended. These findings suggest that public policies can sometimes crowd out private-sector prosocial motives and thus undermine the intended policy effects.

The second paper addresses a methodological question in empirical corporate finance: when and why the outcome variable should be log-transformed in reduced-form regressions. My analysis highlights an important but often overlooked distinction in the economic meaning: using logged outcome estimates average proportional changes, while using non-logged outcome captures average level changes. I demonstrate that different functional specifications can sometimes even produce estimated coefficients with opposite signs. Consequently, I argue that the functional form of the outcome variable should not be solely based on statistical characteristics but also on the research question.

The last paper explores the role of CEO power in uncertain times. Conventionally, regulators and researchers focus on the cost of excessive CEO power. I examine whether powerful CEOs can be beneficial in uncertain times. I measure CEO power by CEO chair duality and uncertainty by industry-level stock volatility. I find that, during uncertain time, powerful CEOs are less likely to be dismissed, and they are associated with better performance. To mitigate endogeneity concerns surrounding CEO power, I exploit the onset of the COVID-19 pandemic as an unexpected, drastic, and sudden surge in uncertain. Firms are unlikely to replace their CEO or change CEO power during such a short period. I find that powerful CEOs are associated with significantly higher stock returns during the one-month period at the beginning of the pandemic. I find suggestive evidence for two mechanisms that contribute to powerful CEOs’ effectiveness during uncertainty: they are more capable of taking swift action, and they are more willing to share information with the board. Overall, this paper challenges the conventional view that CEO power is always manipulative and detrimental.

Overall, my thesis examines the factors that influence corporate decision-making and performance, and discusses a common methodology issue in empirical research.

Authors: Zheng, J

• DOI: 10.5287/ora-xkj67mxrv
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English