Event-inducated Volatility and Abnormal Performance: Shattering the Glass Ceiling: An Impact Study of the Inclusion of Women in the Leadership of S&P 500 Firms

Abstract

This thesis intends to contribute to the study of corporate governance by exploring the effects of gender diversity in leadership positions within the firm and how this affects firm performance. Theory and literature would suggest that gender diversity would have a positive effect on this matter due to the perceived differences between men and women. Differences in such things as risk propensities, behaviors, backgrounds, and qualification requirements can all potentially contribute to a more responsible and knowledgeable leadership team.
These notions will be investigated by exploring gender diversity on the board of directors and executive team to assess whether the inclusion of women in the leadership significantly changes the long-run risk and return characteristics. To this end, two events are defined for this thesis: the inclusion of women on a previously all-male board (PWB) and the inclusion of women on an all-male executive team (PWE). The analysis is conducted using the S&P 500 as a sample over a time horizon ranging from 03.01.2011 to 30.12.2022. A GARCH (1,1)-adjusted market model is selected to simultaneously estimate abnormal return and volatility effects. To test the robustness of results to asymmetric behaviors in the variance structure of returns, a GJR-GARCH (1,1)-adjusted market model chosen upon. Further robustness testing is done by employing a more explanatory five-factor model plus expanding and shortening the event period to infer limits to the long-run effects. We do not find any significant change in the long-run return characteristics for the firms that included women on either the board of directors or executive team. Nor do we find a significant change in the long run volatility for the firms that included women on the board of directors; however, we do show that firms that include women on the executive team significantly increases the long-run volatility and the actual mean of the coefficient for abnormal volatility.
These findings challenge traditional event study methodology that assumes firm-specific conditional volatility of returns to be constant by providing evidence that symmetric and asymmetric behaviors in the variance structure ought to be considered when modelling stock returns.