Banking Markets: Productivity, Risk, and Customer Satisfaction

This paper describes a structural model which incorporates bank decisions on productivity, risk-taking and customer satisfaction into an equilibrium model of banking markets for 219 large U.S. banks, between 1984 and 1992. The cost of bank risk is assessed using banks stock market betas as a broad measure of total risk. In particular, the effect of size on risk-taking is analyzed, as well as the decomposition of risk on a product basis. The Capital Asset Pricing Model is used to measure bank risk, thereby capturing all risk in a measure based on market behavior. Productivity losses due to mismatches between demand capacity are modeled and estimated. In addition, the effect of customer satisfaction, or quality, on bank profitability is measured. Structural model estimation is used to expand the range of questions that can be empirically addressed. It is also used to explain and directly estimate inefficiencies heretofore captured only by inference in fixed effects models. In this structural model banks face long- and short-run cost function choices, they optimally choose capacity levels and risk levels with limited information, and they interact in markets, leading to an asymmetric Cournot-Nash equilibrium. Each bank's profit function is derived from the market equilibrium conditions, and its parameters are directly estimated. The empirical analysis incorporates six bank products and develops a unified approach to thinking about what a bank s products are. The ability of banks to satisfy their customers is examined for the first time in the literature, and it is shown to be a source of profitability A new customer satisfaction data set is introduced to study this phenomenon. All of the relevant factors are integrated into a single model that estimates all parameters simultaneously. The model has five distinct parts: operating activities, risk, demand, customer satisfaction or quality and competitive interactions. The author concludes that banks differ widely in their ability to manage risk. Larger banks take on relatively more risk, with risk cost accounting for 38 percent of bank earnings on average. There also are substantial inefficiencies due to demand/capacity mismatches. On average, banks are over-optimistic by 10 percent in the demand for which they plan, and the cost to them is approximately 2.2 percent of total costs. This is substantially more than can be justified by "optimal overshooting" in the face of planning uncertainty and amounts to over 25 percent of average bank margin. Greater customer satisfaction correlates with greater profitability, principally due to greater demand. The effect of quality on cost and price is minimal. Bank-specific fixed effects are relatively insignificant. The very significant bank-specific effects that previous research discovered appear to have been largely captured and directly estimated in the structural model. The research confirms the results of previous research that there are no significant long-run economies of scale or scope.