Chief Executive Officer (CEO) power, board independence and bank risk in Uganda

Abstract

The study's main aim was to establish the moderating role of board independence in the relationship between CEO power and bank risk. Although risk management has dominated bank management literature and discourse, policy and practice, previous studies on how CEO power affects risk-taking have produced mixed results. Some studies show that CEO power reduces risk, while others indicate that it increases the risk. This lack of conclusive findings is due to the failure to include moderator variables such as board independence that influence that relationship. Failure to include the role of board independence in regulating the extent to which CEO power affects bank risk in the annual reports of commercial banks and the absence of the same in policy documents could also be responsible for bank failures in Uganda over the last 15 years. This study was therefore conducted with four objectives: examining the relationship between CEO power and bank risk, assessing the moderating effect of board independence on the relationship between CEO power and bank risk, analysing the cointegrating relationship between CEO power and bank risk and analysing the causality relationship between CEO power and bank risk. This explanatory panel research used secondary data from a sample of 14 commercial banks in Uganda from 2010 to 2020. The study used secondary data collected through extraction from reports and documents. System General Method of Moments (GMM) was used to establish the relationship between variables and to test the moderating effect of board independence on the relationship between CEO power and bank risk. Diagnostic tests were also carried out to check the suitability of the GMM estimator. Autoregressive Distributed Lag (ARDL) approach was used to infer causality and to analyse the short and long-run linkages between CEO power and bank risk or cointegration. The speed of adjustment of the model in the long run was established using the Error Correction Term (ECT). The findings revealed an inverse relationship between CEO power and bank risk. Commercial banks that have powerful CEOs seemingly have lower risk. Such powerful CEOs have prestige power, are internally hired, have ownership, and have served for more years; hence, they possess expert power. The relationship between current and previous bank risk is positive and significant, confirming a long-run positive relationship between previous and current bank risk. The moderating effect of board independence in iii the relationship between CEO power and bank risk is significant for prestige power and the CEO being internally hired. This means that commercial banks in Uganda should appoint CEOs with outside connections and serve for more years since experienced CEOs reduce bank risk in Uganda. Furthermore, the relationship between their power and bank risk is positively affected by board independence. The findings all confirmed cointegration between CEO power and Z-score in the panel dataset, and that if the model is destabilised and moves away from equilibrium or has short-run disequilibrium, it will correct its previous period’s disequilibrium at a speed of approximately 2.58% annually to get back to the steady state. Lastly, it was determined that there is a causal relationship between CEO power and bank risk. In case there is a need to reduce bank risk in Uganda, adjusting CEO power will help to attain this. It is thus recommended that commercial banks’ annual reports should include the extent to which board independence affected the relationship between CEO power and bank risk. Banks should also encourage CEOs to stay in office for more than four years, to a maximum of seven years. CEOs and bank staff must find new products that will attract people in the informal sector, speed up the readjustment of operations to equilibrium, and reduce risk. CEOs, bank managers, employees and policymakers should not expect immediate results regarding expected changes in bank risk. The results of actions taken in the current year to improve a bank's risk profile can only be seen in the following year. Therefore, there is a need for persistent adjustment and observation of decisions and policy actions if bank risk is to be minimised.