Bank Relationships Earnings Quality and Borrower Default

Abstract

Building on the prior studies on the relationship between earnings quality and bank debt, I investigate the influence of bank monitoring by multiple lenders on accruals-based earning management behavior in an emerging economy, India. I also provide evidence for the heterogeneity in the monitoring function of banks with differences in ownership structure. Consistent with the monitoring hypothesis, I find that increased monitoring by multiple banks deters accruals-based earnings management by borrowers. However, when borrowers add a new bank, I provide evidence that they take advantage of the information disadvantage of the new bank to increase AEM. My results are consistent with the beneficial effect of multiple lenders on bank monitoring to deter accruals-based earnings management. However, the degree of competition among banks, the inability of a large group of lenders to coordinate efficiently, and the time a new bank takes to learn about a new borrower counteract the incentives to monitor a borrower and leads to higher earnings management. In my study on default prediction, I investigate the impact of the number and change in the number of banking relationships (NBR) on the probability of default. The results suggest that NBR is positively related to the probability of distress default, suggesting that multiple banks lower monitoring quality. Opportunistic firms anticipate and exploit coordination failures arising from conflicting objectives among multiple banks. Although banks thoroughly screen the borrowers during loan initiation, they fail to identify borrowers who could defraud them subsequently. I also find that the probability of default increases when the number of public sector banks lending to a firm increase. Further, I find the power of NBRs to predict future defaults strengthened significantly after a regulatory intervention through the Asset Quality Review in 2016. My first essay extends the literature on bank debt and earnings management by providing evidence that although in a cross-section..