Random Utility Models of Demand for the U.S. Commercial Banking Industry

In this paper, we specify a Random Utility Model of Demand for Deposits in the U.S. Banking Industry, assessing its particular characteristics, such as a large number of participants, a large number of markets and an unbalanced panel (many banks participate in only one market and no bank participates in all markets). We modify the standard models to incorporate the fact that deposit balances are different among consumers, in a relationship proportional to their wealth. Using a unique dataset, we estimate the model and find that characteristics other than the interest rate, such as branch density, state presence, etc. add utility to the consumer. The model is also helpful in offering a more realistic set of elasticities among the many banks present in the sample. It shows how market shares will respond depending on the market demographics and current choice set (i.e. offerings of other banks). Finally, we use the results of the model to analyze changes in welfare during the 1994-2002 period. By applying a slightly modified version of Small and Rosen’s equivalent variations, we find that the consolidation process of the late 90s was welfare enhancing, particularly for the middle income consumer.