1. "CEO-subsidiary relationship and banks risk-taking behaviour" with Professor Oleksandr Talavera (University of Birmingham) and Dr. Kim Cuong Ly (Swansea University)
Abstract:
This paper investigates risk-taking behaviour across subsidiaries of U.S. multi-bank holding companies (MBHCs) after a subsidiary-level manager is promoted to the MBHC-level CEO. Our results suggest that CEO-linked bank subsidiaries take more risks compared to their non-linked counterparts. Such an effect is more pronounced when CEOs carry on their employment status with respect to the associated subsidiaries post-turnover. The risk-taking practices at linked banks persist over the years following CEOs succession. Our findings suggest significant merit for regulatory attention on the relationship between CEO and subsidiaries in stabilizing the banking system.
2. "Determinants of CEOs compensation: LASSO approach" with Professor Oleksandr Talavera (University of Birmingham) and Professor Shuxing Yin (University of Sheffield)
Abstract:
Can machine learning assist firm in setting the right level of compensation for chief executive director? We use least absolute shrinkage and selection operator (LASSO), a machine-learning tool, which performs both variable selection and regularization to reveal a subset of factors that are most important in explaining CEO compensation. We find that, out of twenty-five determinants of CEO compensation gathering from literature, a group of seven factors have little or no impact in driving CEO compensation, including CEOs age, gender, firm return on assets, research and development expenses, sales growth, diversification and board size. We emphasize the finding that CEOs gender is not relevant for the CEOs compensation package and the debate of gender pay gap may need a reassessment. We also highlight that accounting measure of firm performance (ROA) should not be tightened with CEOs pay. Overall, the application of machine learning in the area of executive compensation shows potential to help the board of director in the pay-setting process.