Informational Rents in Bank Lending

Bank customers exhibit differing behavior when repaying credits. Lenders are therefore uncertain about borrowersquality. Having superior knowledge about the behavioral type of a customer might generate a surplus for a bank. Crucial in that profit opportunity is the asymmetrically distributed information between the banks. Since it is not obvious which interest rate is paid in such circumstances we will first show the equilibrium conditions and outcomes which are driven by winners curse effects. These will be compared to an information-sharing, symmetric information setting. Banks are indifferent between both settings. A good customer favors a regime which allows banks to share information and for which he has a positive willingness to pay. A bad customer prefers no information sharing.

Banks lend money to borrowers without having any specific knowledge about their repayment behavior. This is part of the risk-taking function of banks. In order to shed light on the credits offered to debtors on such a competitive loan market we conduct an experiment. Each subject represents a bank and is to set repayment claims which have two functions. First, they determine the probability of winning the customer and second, they influence the payoff in case of a repayment. Among other things we find in this experimental study that information acquisition is a motivation for initial lending in general and at moderate rates. If a bank knows a customer well it is able to extract a positive informational rent. Therefore banks tend to attract customers in an initial period with lower offers of repayments and increase their claims independent of information gathered. Letting banks learn borrower type or actual repayment behavior generates similar results.

Banks make use of data about customers to assess the risk of lending money. The information banks receive is viewed as a signal. This signal is not perfectly reliable due to market conditions, and not all banks receive this signal to the same degree of precision. This creates heterogeneity among the banks. The better the banks information the higher is the precision of this signal. Two situations are considered. First, when both banks independently perform a credit worthiness test both receive a private signal. Secondly, when only one of the banks performs a test the signal with higher accuracy is private. This is the case for example when one of bank has e.g. an internal rating and the other bank relies only on a public information source, e.g. an external rating. The better informed bank is in any case able to attain an informational rent which is increasing in the quality of its signal.