December 15, 2006
« More Economic Letters
The Geographic Scope of Small Business Lending: Evidence from the San Francisco Market
Historically, small businesses have tended to turn to local lenders for credit. In recent years, however, technological advances in processing information and assessing credit risk have raised the potential for loosening the geographic ties between small business borrowers and lenders. This Economic Letter discusses factors affecting the geographic scope of markets for small business credit and uses data available for the San Francisco Bay Area to examine the extent to which small businesses rely on local lenders, how this reliance has changed over time, and the implications of any changes for the Federal Reserve’s bank merger policy.
Banking theory provides a rationale for the link between small business lending and local banks. Small businesses typically do not have detailed public financial statements. Therefore, lenders look for other information to determine a borrower’s creditworthiness and to monitor existing loans. Historically, much of that information was most efficiently gathered locally, and it might include knowledge of the local economy and community, knowledge of the borrower’s industry (since some industries are concentrated locally or regionally), knowledge of the borrower’s particular firm, and knowledge of the “character” and skills of the small business owner or owners. Brevoort and Hannan (2004) find direct empirical evidence that lender and borrower location do matter in small business lending, even within local areas. Using data for local banks and local businesses, these authors find that banks lend more to nearby borrowers than to distant borrowers, other factors held equal.
Among local banks, small banks may be particularly well-suited to small business lending. For example, Berger, Klapper, and Udell (2001) argue that small banks have a comparative advantage over large banks in making “relationship” loans, that is, loans that use “soft” information. Soft information is more qualitative and subjective than “hard” information and may be relatively difficult to communicate. Among the types of local information already mentioned, information regarding the character of the business owner, a key element in many small business lending decisions, seems particularly open to interpretation. In theory, both large local banks and small local banks can obtain character information rather easily, for example, through repeated contacts between the loan officer and the borrower in either informal social settings or face-to-face business meetings. However, loan officers at small banks may be able to use that information more easily than those at large banks, because they are likely to be able to convey their impressions through fewer layers of management. Indeed, in 2005, small (less than $1 billion in assets) local banks in the San Francisco Bay Area (the San Francisco-Oakland-Fremont Metropolitan Statistical Area) are estimated to have held 6.6% of their Bay Area assets in Bay Area small business loans, versus only 0.6% for large banks with offices in the Bay Area.
At the same time, there are reasons to expect that technological advances in data-gathering and risk-assessment are working to reduce the importance of geographical proximity between lenders and their small business customers. One example is the prevalence of credit scoring, in which computer-generated models use a limited number of factors, including an applicant’s loan repayment history, to determine the probability that a borrower will default on a loan. Lenders have long used these models to assess individuals’ creditworthiness, and they now also use them to qualify some small businesses for loans. Scoring models in essence automate the credit underwriting process and can limit the added value of more qualitative information that might be gleaned through so-called relationship lending by a local lender. If so, this would tend to broaden the geographic scope of the market for credit at least for some small business borrowers, thus making non-local lenders more effective competitors in small business lending markets.
To examine the extent to which small businesses turn to local lenders and how that tendency has changed in recent years, I examine small business lending data for the San Francisco Bay Area. Part of the data comes from reports that large banks and all banks in large bank holding companies are required to file with their regulators under the Community Reinvestment Act (CRA). In using these data, small business loans are defined as term loans and lines of credit no greater than $1 million made to businesses with gross annual revenues no greater than $1 million. During the period under study, 1998 to 2005, banks’ CRA reporting requirements changed, and I was able to obtain information that allowed me to correct for the change in reporting in the Bay Area. Such information was not available for other areas. I estimate the small business lending of Bay Area banks that did not report on the CRA with data from the Reports of Condition and Income and from the Summary of Deposits.
Two criteria are used to group the banks: (1) whether the entity is a “credit card bank” or just a “bank,” and (2) whether the entity is local (its own or an affiliates’ offices in the Bay Area) or “out-of-market” (non-local). Credit card banks are defined as commercial banks with the bulk of their business loans in the form of business credit cards or those with personal credit card lending as a main activity (at least half of assets in consumer loans and at least 90% of consumer loans in credit card loans). The category “bank” includes other commercial banks, savings banks, savings and loans, and industrial loan banks. (Note that other potentially important sources of small business credit, such as finance companies, venture funds, “angel” investors, customers, suppliers, friends, family, and personal credit cards, including those held by small business owners, are excluded from the analysis.)
Credit card banks are considered separately for two reasons. First, innovations in underwriting, particularly the use of credit scoring, have played a major role at credit card banks. Second, in analyzing the competitive effects of bank mergers on small business lending in local banking markets, the Federal Reserve distinguishes between credit card banks and other depository institutions.
When applied to the Bay Area, three groupings of lenders are applicable: (1) local banks, (2) out-of-market banks, and (3) national credit card banks. Below, small business loans from credit card banks will be synonymous with credit card loans to small businesses, and small business loans from other banks will be synonymous with non-credit card loans to small businesses. In practice, of course, credit card banks make some non-credit card loans, and other banks make credit card loans.
The estimates for the Bay Area show that, in 2005, local banks held most of the dollar volume of small business loans (84.6%), while national credit card banks held most of the remainder (14%) (Figure 1). The negligible share held by out-of-market banks is not surprising, given evidence from the 2003 Survey of Small Business Finances, which indicates that the median distance between small businesses and their non-credit card depository institution lenders is only four miles.
The most recent data, then, show that local banks dominate Bay Area small business lending. Is there evidence that the geographic scope of the market for small business credit is expanding? As Figure 1 shows, the market share of local banks did decline somewhat between 1998 and 2005—from 91% to 84.6%. The share of out-of-market banks also shrank.
The lenders gaining share clearly are the credit card banks. Their cut of the pie more than tripled, from 4.5% in 1998 to 14% in 2005. Since credit card banks have widely adopted credit scoring, their gain in small business lending is at least consistent with innovation working to loosen the geographic tie between small business borrowers and lenders.
The growth of credit card small business lending could have implications for the analysis of the effects of proposed bank mergers on competition in small business lending. When reviewing bank merger proposals, the Federal Reserve analyzes the proposal’s potential competitive effects using each bank’s deposits in the local banking market to measure its share of that market. Local branch deposits proxy for the bank’s supply of banking services as a whole, including small business loans. When this broad analysis indicates the possibility of significantly large anticompetitive effects resulting from the proposed merger, the Federal Reserve often also analyzes the potential competitive effects of the merger on small business lending in particular, using loans made to small businesses in the local market to measure market share.
Under current practice, small business loans made by credit card banks are excluded when calculating banks’ small business lending market shares. Insofar as the data for the Bay Area may be representative of trends elsewhere in the country, credit card banks’ recent gains may argue against such exclusion.
However, in judging the relevance of credit card lending, we need to consider the degree of substitutability between small business credit card and non-credit card loans. For example, as with personal credit cards, business credit cards have relatively low loan limits, so credit card loans tend to be much smaller than local non-credit card bank loans. For example, in 2005, the average size of a local (non-credit card) bank small business loan was $55,468, versus only $8,756 for a credit card bank small business loan. After all, credit scores can yield only so much information and cannot, for example, indicate how business owners respond to uncertainty or how the business fares over economic cycles. With a much richer set of information, local non-credit card banks can likely better manage the risk entailed in lending relatively large sums than can credit card banks. And, since credit card loans are smaller, they tend to be used for working capital purposes, unlike the capital investment expenditures for which local non-credit card small business loans are more likely to be used.
Moreover, in at least some cases, business credit card contracts prohibit small businesses from holding multiple credit cards from multiple banks. In such cases, small businesses cannot circumvent credit limits on their cards by borrowing against multiple cards and, therefore, cannot duplicate a larger non-credit card loan with multiple smaller credit card loans. In addition, again as with personal credit card loans, interest rates on credit card loans tend to be much higher than for non-credit card loans, for similar terms to maturity. Therefore, small business borrowers’ incentives to replace non-credit card loans with credit card loans may be limited.
Small businesses have tended to rely on local lenders for credit. The dominant role of local lenders often is ascribed to information advantages associated with having a presence in the region in which the business customer operates. Data for the San Francisco Bay Area, however, show that in recent years national credit card banks have gained some market share in small business lending. The gains are consistent with expectations that advances in information technology, risk assessment, and risk management should have helped loosen the geographic ties between small business borrowers and lenders.
Nevertheless, those ties appear to be far from severed. The most recent data still show that banks with their own or affiliates’ offices in the San Francisco Bay Area account for about 85% of the lending to small business by traditional intermediaries. Moreover, out-of-market lenders other than national credit card banks have lost market share in the Bay Area, suggesting that distance still is a deterrent for some lenders. And the relatively small average amount of credit extended via credit cards raises some question about the likelihood that credit card lenders ever will substitute fully for local lenders.
Berger, Allen, Leora Klapper, and Gregory Udell. 2001. “The Ability of Banks to Lend to Informationally Opaque Small Businesses.” Journal of Banking and Finance 25(12), pp. 2127-2167.
Brevoort, Kenneth, and Timothy Hannan. 2004. “Commercial Lending and Distance: Evidence from Community Reinvestment Act Data.” Finance and Economics Discussion Series 2004-24, Board of Governors of the Federal Reserve System.
More Economic Letters
Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Sam Zuckerman and Anita Todd. Permission to reprint must be obtained in writing.
Please send editorial comments and requests for reprint permission to