The use of accounting information in bank lending decisions

This paper examines the impact of accounting information on the sequential judgments of experienced bank loan officers using realistic lending cases in an experimental setting. The findings suggest that loan officers reach a high level of confidence early in the lending process based on summarized accounting information and other general background data. When, later in the process, factors concerning the firm's financial plans and their underlying assumptions are varied, lenders adjust their confidence in whether or not to grant the loan in the expected directions, even when the subsequent evidence disconfirms their original positions. Accounting information is used in many business decisions, including the very important area of bank lending. Even though financial accounting is said to be developed to assist external users in their business decisions, with the two primary external user groups identified as investors and creditors (FASB, 1978), there is very little empirical work that examines lending decisions and how creditors process accounting data. We examine the impact of accounting information on the sequential judgments of bank loan officers using realistic condensed lending cases in an experimental setting. The study reflects a high degree of external validity, and provides new insights into how and when various types of accounting information are used in banks' credit-granting decision process. The results of our research suggest that lenders reach a high level of confidence in their credit-granting decisions very early in the information gathering and evaluation process. This initial confidence is based on summarized financial accounting data and other background information. At the same time, lenders' confidence in their judgments at subsequent stages of the information gathering and evaluation process are materially altered by more detailed forms of accounting data. Even in cases where subsequent information disconfirms prior judgments, lenders respond by revising their confidence in the credit-granting decision consistent with the nature of the subsequent evidence. While this result is consistent with the common finding that strong initial confidence is obtained early in the decision process, it is inconsistent with the many studies that find disconfirming evidence not to be searched for and/or to be ignored once a strong initial impression has been formed. The study provides evidence that accounting data have a material impact on lending decisions, and that summarized basic financial statement data, collected independently of the client borrower early in the process, have the greatest impact on their decisions for the class of borrower examined in this study. These results point to a previously unexplored source of data for future investigation into accounting's role in the creditgranting process. The next section of the paper discusses the nature of the credit-granting process, and the subset of credit-granting decisions examined in this study. The third section of the paper *The authors appreciate the financial support of the Touche Ross Foundation, which made this study possible. We appreciate the assistance of Carol Frost, Jaysri Sankaran and Jacob Thomas.