Academic journal article Quarterly Journal of Finance and Accounting

The Importance of Being Known: Relationship Banking and Credit Limits

Academic journal article Quarterly Journal of Finance and Accounting

The Importance of Being Known: Relationship Banking and Credit Limits

Article excerpt

Introduction

Previous work on the National Survey of Small Business Finance (NSSBF) has examined the effect of bank-firm relationship durations on the interest charged (Petersen and Rajan, 1994; Berger and Udell, 1995), on the probability of the loan being secured by collateral (Berger and Udell, 1995), and on the probability of the loan application being rejected (Cole, 1998). This paper examines the effect of duration of the bank-firm relationship on the allocation of credit for a sub-sample of firms in the NSSBF. More specifically, we show how bank relationships affect the size of credit limits (CL)--the maximum amount of loan that can be drawn under a line of credit (LOC) agreement. Unlike earlier work, we treat a credit contract as a package of loan terms, where the interest rate, collateral requirements, and the size of the loan are simultaneously determined. This approach combines the literature on relationship banking with the literature that views a credit contract as a package of loan terms allowing for trade-offs among its various parameters (as in Azzi and Cox, 1976; Arzac, Schwartz, and Whitcomb, 1981; Koskela, 1983; Melnik and Plaut, 1986).

A bank's ability to produce reliable private information lowers the information asymmetry between bank and firm and facilitates credit contracts. [1] The embeddedness approach in sociological theory and anecdotal evidence suggest that bank-firm interactions over time lead to private networks of social nexus between entrepreneurs and bank officials facilitating credit (Uzzi, 1999; Degryse and Ongena, 2001; Ongena and Smith, 2001; Lehmann and Neuberger, 2001; Machauer and Weber, 1998; Agarwal and Elston, 2001). [2] Access to such private information and networks, or relationship banking, is particularly important in lending to informationally opaque, personality-driven small businesses with limited access to equity markets (Petersen and Rajan, 1994). As Stein (2002) points out, credit to small business is often character loans, based on soft information about the entrepreneur revealed over time. Hence, the duration of bank-firm relationships has been used as a proxy for this evolving information. The straightforward strategy of the NSSBF-based empirical literature has been to use as the dependent variable either (a) the interest rate, (b) the binary variable on whether the loan was secured with collateral or not, or (c) the binary variable on whether the loan application was rejected or not in regression analysis and then use the bank-firm relationship duration as the independent variable while controlling for firm, bank, and owner characteristics.

By looking at the difference in the credit limits that two banks have provided the same firm on line of credit facilities, our econometric specification eliminates the need to control for owner and firm characteristics including firm age. Firm and owner characteristics are treated as fixed effects that both contracting banks observe uniformly. The duration of the bank-firm relationship as a proxy for private information has its weaknesses such as high correlation with firm age, which is a proxy for public information. Cole (1998) argues for the importance of disentangling these variables and presents this as the main reason for Petersen and Rajan's (1994) inability to show a clear relationship between the risk premium on the most recent loans and bank-firm relationship durations. [3]

The differencing method we use also allows us to use instrumental variable (IV) estimation techniques to adjust for the simultaneity in determining a package of loan terms by freeing the firm and owner characteristics to be used as instruments. Thus, this paper differs from the literature by being able to include more than one set of loan terms in the equation being estimated. We find that bank-firm relationship durations are significant determinants of credit limits under instrumental variable estimation for firms with multiple credit lines. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.