Dissertations and Theses @ UNI

Availability

Open Access Thesis

Keywords

Banks and banking--Iowa; Competition; Banks and banking; Competition; Iowa;

Abstract

This study presents an analysis of the competitive process in banking in Iowa. The analysis is accomplished by developing and testing an extension of a statistical model originally introduced by Bernard Shull. This model describes the commercial bank as a multiple-product, price-discriminating firm set in an environment of monopolistic competition. The model is estimated across time using pooled data for the years 1967 through 1971. No previous study on this subject has attempted the use of pooled time series and cross sectional data. Pooling the data necessitates that the explanatory variables be examined as Xij such that X represents the observation of the ith individual object in the jth period of time. Thus, the resulting dependent variable represents the jth observation of the ith individual object. This statistical method permits an examination of banking performance over time. Because of the regulated environment in which banks operate, and because of structural changes inherent to the banking industry, the dynamism provided by this statistical methodology is considered of major importance to the study. The testing of the model is accomplished by a stepwise regression technique which entails forcing certain variables into the regression specification while allowing others to enter freely. From this testing, it was hoped that the model would furnish an elasticity estimate of the overall market performance of the commerical [sic] bank. From the overall elasticity estimate the elasticity measures for eight individual subproduct loan markets could be derived. These classes are: residential real estate loans; other real estate loans; commercial and industrial loans; loans for the purchase and carrying of securities; agricultural loans; loans for automobiles; all other loans; and other loans to individuals and loan to individuals for credit cards. These loan classes represent the primary loan mix model and are the variables that are forced into the equation. The variables which are left free to enter are: banking operating structure; market structure; geographic area; and banking acitivity [sic]. In addition, a yearly dummy variable scheme is added to account for structural changes from year to year. A major difficulty encountered in testing the statistical model is that of obtaining statistically significant results. This problem diminishes somewhat if the data for 1969 is omitted from the study. However, on the basis that pooling time series and cross sectional data is a beneficial technique, it would be fallacious to delete an entire year from a set of time series observations. When each year is tested independently, the model proves not only explanatory, but also predictive. Although pooling the data does yield a satisfactory R2 value, and hence might be construed as a feasible predictor, the lack of statistical significance of the regression coefficients disallows any structural interpretation to be given to the statistical results. Possible causes for these negative findings are: (1) structural changes from year to year are so great that they cannot be adjusted for; (2) because of the aggregate nature of the data, the resulting interaggregate bias causes the model to be autoregressive.

Year of Submission

1974

Degree Name

Master of Arts

Department

Department of Business

First Advisor

Marion Chiattello

Second Advisor

Thomas Reuschling

Third Advisor

B. Wylie Anderson

Comments

If you are the rightful copyright holder of this thesis and wish to have it removed from the Open Access Collection, please submit a request to scholarworks@uni.edu and include clear identification of the work, preferably with URL.

Date Original

1974

Object Description

1 PDF file (44 leaves)

Language

en

File Format

application/pdf

Download

Comments?

Included in

Business Commons

Share

COinS