Behavioral factors in banking as determinants of the interest rate spread.

The aim of this paper is to explore behavioral factors influencing bank clients and to quantify their effect on profitability as measured by the interest rate spread. For this purpose we use a traditional model, which includes common macroeconomic, financial structure and institutional indicators from the existing literature as control variables along with a number of behavioral variables combined in an index. These include the number of bank branches per person, the number of ATMs per person, percentage of people with tertiary education, and market capitalization of listed companies as a percentage of GDP. The behavioral index aims to capture the tendency of bank clients to make decisions, which are not solely based on the financial parameters of the bank’s offer. We use panel data on twenty two European countries, spanning the eleven year period from 2003 till 2013 and employ a system GMM estimation. The results show that the behavioral index influences the interest rate spread positively. On the other hand, we cannot draw conclusions about banks’ overall profitability as net income is also affected by the level of noninterest income and incurred costs, which are likely to be higher in countries where less rational behavior is present and banks have to tend to the clients’ non-financial preferences. Consequently, when clients are acting in a less rational manner, i.e. having low involvement or low confidence, financial institutions build larger networks, which increases their costs. Since we observe that the interest rate spread is larger in countries where less rational behavior is present, these costs seem to be passed on to the clients. In countries where customers seem to be more concerned about the financial aspects (i.e. act more rationally), on the other hand, financial institutions are forced to compete on price and therefore charge lower spreads. In addition, we demonstrate that the inclusion of a behavioral index improves standard models for the estimation of the interest rate spread.