07/09/2024
By Cuibing Wu
Candidate Name: Cuibing Wu
Defense Date: Monday, July 22nd, 2024
Time: 1-3 p.m.
Location: Virtual - Zoom link
Thesis/Dissertation Title: Three Essays on Analyzing the Impact of CEO Interview Videos on Firm Performance.
Committee Members
1) Julie Zhang (Co-chair), Ph.D., Department of Operations & Information Systems, Manning School of Business, UMass Lowell
2) Xiao-bai (Bob) Li (Co-chair), Ph.D., Department of Operations & Information Systems, Manning School of Business, UMass Lowell
3) Shakil Quayes, Ph.D., Department of Economics, College of Fine Arts Humanities & Social Sciences, UMass Lowell
Abstract:
Chief Executive Officers (CEOs) as the companies' leaders and key decision-makers, can provide important information that may be useful in predicting the organization's future course and general success. This dissertation focuses on the multi-modal cues from CEO interview videos to determine whether and how CEO interviews impact firm performance. Through three interconnected studies, we provide a comprehensive understanding of the multi-modal cues from CEO interview videos.
In the first essay, we examine the influence of CEO interview videos on firm market value. We provide statistically significant evidence that CEO interview videos have a positive effect on a firm's market value by employing a staggered difference-in-differences (SDID) model. The findings are supported by both a price model and a two-stage treatment effects model used in existing studies on information disclosure and firm values.
The second essay investigates the impact of verbal and nonverbal cues in CEO interview videos on market volatility from two facets. We first use advanced machine learning techniques to quantify verbal, vocal, and visual cues to examine their relationship with market volatility. Then, in the experimental phase, Generative Adversarial Networks (GANs) are used to generate varying emotional expressions in CEO videos, and participants are shown these videos. This comprehensive study highlights the significant influence of CEO verbal and nonverbal cues on market volatility.
The third essay introduces the Hierarchical Cross-Attention Transformer (HICAT) model to analyze CEO interviews using multi-modal features and inter-modal inconsistency. Our findings demonstrate that HICAT significantly improves predictive accuracy in forecasting financial risk compared to traditional baseline models. Additionally, integrating HICAT with a fundamental model through a stacking method further enhances predictive performance. Our work provides deeper insights into how CEOs' verbal, vocal, and visual cues interact and influence financial risk, offering a robust framework for investment decision-making.
All interested students and faculty members are invited to attend.