Some Determinants Of Commercial Bank Behavior

The central theme of this dissertation is commercial bank behavior. Following the introduction in Chapter 1, Chapter 2 examines U.S. banks' choices of foreign activities. Chapter 3 analyzes Hungarian commercial banks' branch network and interest rate choices. Specifically, Chapter 2 relies on a theoretical model and estimation to examine how banks' scope of operations and size, and various host market characteristics, determine banks' choices of foreign market entry/exit, and foreign loan/deposit quantities. Applying the Bajari, Benkard, and Levin (2007) two-step estimation method, the determinants of the optimal foreign loan and deposit choices are estimated in the first stage. The results (1) confirm the presence of and correct for significant selection bias arising from the correlation in banks' entry and loan volume choices; (2) show different sensitivities of cross-border and affiliate loans to market and bank traits, and (3) characterize the role of bank scope in bank behavior. In the second stage, forward simulation is used to estimate banks' and regulators' risk aversion parameters, the fixed foreign market entry costs and scrap (liquidation) values. Results show that entry costs are higher in inefficient and profitable markets with greater entry barriers and stronger government presence in banking. Scrap values move together with entry costs, and regulators are more risk averse than banks. Regulatory risk aversion is greater in inefficient markets with stricter regulations and lower enforcement power. The chapter concludes with simulation exercises that describe the strong discouraging impact of regulations on U.S. banks' foreign participation. Chapter 3 examines the dynamic behavior of imperfectly competitive Hungarian banks. The chapter consists of a theoretical model and empirical analysis. A simulation-based estimation method is applied to bank-level data to estimate the impact of bank and market traits on optimal interest rate and branch network expansion choices. Estimation results confirm the importance of branch network competition. Furthermore, branch setup cost and scrap value estimates are high, and strongly positively correlated with each other and with relevant producer price indices. Chapter 3 concludes with various simulation exercises, finding a strong impact of competitors' branch network size on bank behavior