Date of Award


Document Type

Open Access Dissertation


Moore School of Business


Business Administration

First Advisor

Allen N. Berger


In my first essay, using a novel dataset that merges the Dealscan database and 8-Ks between 1994-2014, I find that on average, only about 31% of bank loans are announced by firms. Among those loans announced, about 60% are cleanly announced and 40% are announced together with other events. The three-day Cumulative Abnormal Stock Return (CAR) following a loan announcement is positive and both statistically and economically significant, which on average about +39 b.p., in line with the theory of bank loan specialness. This finding is mainly driven by bank-dependent firms. Next, I find significant evidence of sample selection bias in the loan announcements sample, which likely confounds the findings from the previous literature. Correcting for the bias, I find that a loan is more likely to be announced during a market crisis, relative to normal times, but not during a banking crisis. Moreover, a loan is more likely to be announced by small firms, firms with lower EBITDA, and when the loan has more financial covenants, is a revolver loan, has a longer maturity, secured, and when the firm has a previous lending relationship. Then, CARs are significantly higher during a banking crisis, compared to normal times, but not in a market crisis, in line with both the asymmetric information hypothesis and the institutional memory hypothesis. Lastly, I find strong evidence that CARs are negatively associated with the market share of nonbank lenders, which aligns with the competition hypothesis between bank and nonbank lenders.

In my second essay, using a natural experiment of changes in deposit insurance deposit insurance coverage limit over 2002-2011 in Indonesia, I find a significant positive relation between explicit deposit insurance coverage and bank risk-taking, consistent with the moral hazard hypothesis. More specifically, controlling for various bank-specific and macroeconomic variables, as well as bank regulations, I find that Indonesian banks’ Z-Score, an inverse measure of bank risk taking, increases on average about 18% when the government switched from the blanket guarantee era to the limited guarantee era. Further, I find some evidence that the relation is non-monotonic at the low level of explicit deposit insurance coverage, in line with the safety net hypothesis. Finally, I find significant evidence that the impact of explicit deposit insurance coverage on bank risk is different across different kinds of ultimate owners. In particular, family banks and politically connected banks are those that are most affected when the government switched from the blanket guarantee era to the limited guarantee era, suggesting that the moral hazard problem in these banks are more prominent compared to foreign banks and nonpolitically connected banks.

In my third essay (co-authored with Allen N. Berger, Sadok El Ghoul, and Omrane Guedhami), we examine the impact of geographic deregulation on bank risk. More specifically, we study all three types of geographic deregulation in last three decades in the U.S. banking industry—intrastate branching, interstate banking, and interstate branching. These deregulations provide unique empirical settings to test the impact of competition and diversification on bank risk. We find statistically and economically significant evidence that on average, interstate banking deregulation is associated with about 22% increase in Z-score, an inverse indicator of overall bank risk.

On the contrary, we find some evidence that intrastate branching is associated with a decrease in Z-score about 3%. Meanwhile, we find no evidence that interstate branching affects bank risk. These findings are robust to a variety of sensitivity checks, including those for endogeneity and sample selection bias, as well as alternative risk measures. Different than most of the previous studies that focus on large banks and Bank Holding Companies, our findings show that the favorable impact of interstate banking deregulation on bank risk are driven by small banks, which had opposed the deregulation with the fear that an increase in competition from large banks could reduce their survival probability. Meanwhile, intrastate branching is associated with higher risk for small and medium banks, but lower risk for large banks. These findings suggest that the competition-stability channel dominates for small and medium banks, while the diversification-stability channel dominates for large banks.