A MODEL OF BANK BEHAVIOR UNDER UNCERTAINTY: AN EXPANSION AND TEST OF THE MASON MODEL

Presently, there exists numerous microeconomic models of commercial banks. These numerous models can be classified as one of the following: models that require monopoly power for a solution, models that utilize a portfolio structure, or models that recognize the resources used in producing banking services. Unfortunately, no narrowing of focus in the theoretical work has occurred. One reason for numerous models and the lack of focus is the lack of empirical tests of bank models. In the literature, there have been two empirical tests of theoretical bank models: Parkin (1970) and Graddy and Kyle (1979). Parkin\u27s test of his highly detailed model of English discount houses supported a portfolio structure, but the results should not be generalized to a U.S. commercial bank. Graddy and Kyle test a very unstructured bank model to demonstrate the interdependence of asset and liability decisions. This research empirically tests the model published by John Mason (1979). The Mason model was chosen because it modelled bank management as a price-discriminating, risk-neutral monopolist facing a portfolio decision. It also provided the structure for a test of a certainty vs. uncertainty model of bank behavior. For the purpose of this research, the Mason model was modified to include the institutional constraints of reserve requirements and interest rate ceilings. In addition, a proof of a unique equilibrium was presented and the comparative statics were examined. The Mason model was tested by fitting balance sheet and interest rate data from a single bank holding company to the mathematical form implied by the first-order conditions of the model\u27s solution. The OLS regressions were fitted for three loan and five liability categories. The regression results provide (1) an endorsement that uncertainty models are a significant improvement over certainty models, (2) limited support for the Mason model in five of the eight regressions, and (3) evidence that large U.S. banks acquire the majority of their assets and liabilities in competitive markets. This research suggests that banks should be modelled as competitive firms that operate in a world of uncertainty