Ron J. Feldman - Senior Financial Specialist
Published April 1, 1997 | April 1997 issue
Personal contact between small business owners and their banks is at the heart of community commerce in towns and rural areas throughout the Ninth Federal Reserve District. The connection between community banks and small business is not merely the result of the "everyone-knows-everyone" environment that often characterizes nonmetropolitan life. Many bankers will not extend credit unless they have in-depth knowledge of the plans and operations of the prospective small business borrower. However, what if knowledge of local conditions and face-to-face contact with the borrower is unnecessary for making sound loans? Indeed, what if gathering such information actually hurts the competitive position of the community bank? The recent but growing ability of lenders to underwrite small business loans via a highly automated and cost-effective process known as credit scoring moves such questions from the hypothetical to the practical.
The move to credit scoring could have broad implications for small banks, small businesses and small towns despite the seemingly technical or arcane nature of this change. Credit scoring should greatly reduce the time and paper needed to apply forand monitora small business loan. These systems should also decrease the costs of borrowing for small businesses with higher credit quality, and conversely, increase the costs for borrowers with lower credit quality, while increasing the total amount of funds available for small business lending. In addition, credit scoring should put downward pressure on the profits that community banks currently earn on small business loans, although it may open up new areas of profitable business. Finally, credit scoring could alter the environment in which community commerce takes place. Small business lending is often characterized by intense, personal contact between the lender and borrower. Credit scoring makes it much more likely that this relationship will become like the impersonal tie between consumers and their credit card issuers.
This article provides an in-depth look at credit scoring, detailing what it is, how it is used to evaluate potential small business loans, the implications of credit scoring for small business borrowers as well as the implications of credit scoring small business borrowers for community banks.
Credit scoring is the process of assigning a single quantitative measure, or score, to a potential borrower that reflects the borrower's relative chance of going into delinquency or default.
These scores have been used for decades in the underwriting of consumer loans, such as credit card and auto loans.
There are two main types of credit scores: the bureau or credit history score and the application score.
The bureau or credit history score is based solely on information available from a credit bureau. In developing the bureau scoring system, analysts identify those characteristics of borrowers that best predict whether they will repay their loans fully and on time. The developer then statistically assigns each of the characteristics a numerical weight so that the credit system correctly measures the relative importance of a given characteristic in accurately forecasting repayment. For example, the characteristic "never having been delinquent" will receive a large weight if it is an important predictor of not being delinquent in the future. The outcome of these statistical procedures is a "score card" on which the characteristics of the borrower are noted and the score tallied. Fair, Isaac and Co., San Rafael, Calif., in conjunction with each of the three major credit bureaus, is a leading provider of bureau scores. Their bureau scores range from about 300 to about 900. The higher the score, the more likely the loan will perform according to the expectations of the lender.
The second type of score, the application score, incorporates information on the loan and borrower beyond that collected by a credit bureau. A mortgage score, for example, could be influenced by the amount of the loan relative to the value of the property being financed.
Credit-scoring systems do not approve or reject a loan application. Rather, the underwriter must decide how he or she will incorporate the credit score into the loan review. Freddie Mac, one of the two enormous government-sponsored purchasers of mortgages in the United States, has suggested that lenders subject mortgage applicants to three levels of scrutiny based on the bureau score of the applicant. Freddie Mac suggests that those applicants with a Fair, Isaac score of more than 660 should receive a limited review by human underwriters, those with scores between 660 and 620 should receive a full-scale human underwriting while those with scores under 620 should be approved only with great caution. Ultimately, each lender must set its own cut-off scores based on its risk preferences.
The recent but growing ability of lenders to underwrite small business loans via credit scoring moves a string of important questions on loan administration from the hypothetical to the practical.
The popularity of credit-scoring systems results from the advantages that they offer the borrower and lender. Credit scoring has allowed consumer lenders to:
Credit scoring historically had been limited to the underwriting of consumer loans. In contrast, lenders believed that they had to assess the ability of the small business to generate revenue to determine its repayment prospects. Such a comprehensive business review is not amenable to credit-scoring systems. More recently, analysts determined that the bureau score of the owner of the small business is highly predictive of the loan repayment prospects of the business. The credit score of the business principal can be augmented with very basic information on the nature of the business (e.g., is it an entertainment provider or a manufacturer?) and data from a business credit bureau, such as Dun & Bradstreet, to produce one type of credit-scoring system for small business loans.
The most prominent small business loan scoring system was developed by Fair, Isaac and the Robert Morris Association (RMA). It provides lenders with a number of scorecards whereby they can choose the system most appropriate for the characteristics of the loans they are offering. For example, scorecards have been designed for different sized loans and for businesses that do not provide lenders with financial statements. Fair, Isaac reports that 250 banks use its small business scoring system, while 25 of the top 30 banks with the most outstanding small business loans credit score their own small business loans. Several extremely large banks, including Wells Fargo, Citibank, and Bank of America, also have developed their own scoring systems. Wells Fargo alone issued $1.4 billion of preapproved lines of credit to small businesses in 1995.
Credit scoring will alter small business lending in three areas: the interaction between borrowers and lenders, the pricing of loans and the availability of credit. More generally, all of the characteristics that credit scoring has brought to consumer loans will now be available for small business loans.
Traditionally, a small business owner wanting a loan had to meet with a loan officer of a local bank in person and submit an application, including financial statements, business plans and a variety of other records. It could take many days for the loan officer to review the documents. And not just data were examined during the loan review; the "character" of the borrower could play a significant role in the decision. After the loan was made, the bank required that the small business owner regularly submit updated financial statements. Credit scoring eliminates the need for this level of interaction. In fact, by using a credit-scoring system, a lender with no physical presence in a community can lend money to small businesses via a direct solicitation without ever seeing a business plan or financial statements. Once approved, the borrowers may not have to provide their lenders with updated financial statements. If a small business owner comes to the branch of a bank that credit scores its loans, his or her application can be approved very quickly. The credit-scored small business loan would have much more in common with the credit card loan obtained by the business owner in terms of borrower-lender interaction than the traditionally underwritten small business loan.
The price of small business loans will decline for higher-credit-quality borrowers under a credit-scoring regime because these borrowers no longer have to bear the cost of a full human underwriting. Moreover, these high-quality borrowers will now have access to a greater number of lenders. Lenders from across the country will be able to reach out to the small business via direct marketing. This increase in competition should also reduce the costs of funds to small businesses. Finally, some businesses that had previously been thought to be high risk under a traditional underwriting system may be classified as lower risk under a credit-scoring system. This reclassification would encourage lenders to reduce costs for such small business borrowers.
Not all borrowers, however, will see their loan costs decrease because of credit scoring. A borrower whose credit score indicates that a full-scale human review is required may actually face higher costs. Previously, the fixed costs of human underwriting were spread among all the applicants. Under credit scoring, a significant percentage of the loans will receive a limited human review, thus forcing those reduced numbers of loans that still require a full-scale human review to bear the bulk of the fixed costs of traditional underwriting. Furthermore, and as noted above, credit scoring makes pricing according to risk much more feasible. The small-business borrowers who were being undercharged based on the traditional underwriting regime relative to the risk of default or delinquency they posed will now face higher costs for credit. Using its credit-scoring system, Wells Fargo, for example, has recently been charging its higher-risk borrowers the prime rate (about 8.5 percent as of the end of March 1997) plus an additional 8.75 percentage points.
Better information about the repayment prospects of a small business applicant makes it more likely that a lender will price the loan according to its expected risk instead of denying the loan out of fear of charging too little and losing money. This prospect should increase the availability of credit to small businesses.
More important is the indirect ability of credit scoring to increase the size of the pool of funds small businesses can tap. Currently, it is very difficult for a lender that has made small business loans to sell the loans to investors. In contrast, lenders routinely sell their mortgages and, to a lesser extent, their credit card loans. The difference between the two situations is that a large market exists for securities backed by a group of mortgages, credit cards or other consumer loans. In this market, the investor will buy a security whose principal and interest payments are actually supported by payments made on the underlying mortgages or credit card loans. The issuer of these asset-backed securities will use the funds of investors to make additional mortgage or credit card loans. The process of converting a pool of loans into a security that investors are willing to purchase is aptly called securitization.
There is a very small market for securitized small business loans for two reasons. First, there is not much data on how these loans perform over time. And second, small business loanseven those made by the same bankmay vary in their underwriting, payment terms and loan structure. Both these factors make it very difficult for an investor who wants to buy a security backed by a pool of these loans to determine the cash flows that such a pool will produce. Investors are unlikely to buy securities backed by an uncertain cash flow under conditions amenable to the firm that originated the loans. The investors may ask the lender, for example, to sell the loans for less than what the lender thinks they are worth. Credit scoring small business loans addresses both these problems. The highly computerized scoring systems make it easy to collect data on the performance of loans over time. To use a credit-scoring system cost-effectively, a lender must also make its small business loans fairly homogenous. Otherwise, the system will not be able to process many loans in a short period. Using a scoring system to rate heterogeneous loans would be like using the same machine to process many differently shaped and sized widgets. In total, the credit-scored small business loans should be much easier to securitize.
A vast pool of funds opens up to small businesses once lenders can securitize small business loans. Indeed, investors throughout the world who currently invest in asset-backed securities would be able to invest in small business loans.
Credit scoring should increase the competition for lending to small businesses in several ways. Large banks are already applying direct distribution techniques to lend to small businesses that previously borrowed from local institutions. More generally, a firm capable of raising funds and buying a scoring system can more easily enter the small business lending market. Furthermore, lenders making use of credit scoring can provide credit to high-quality borrowers at a lower rate than can banks that rely on the traditional underwriting technique. Finally, the securitization of small business loans would greatly increase the number of potential investors in these loans. All these factors should put competitive pressure on the prices that community banks can charge small businesses. Ultimately, this pressure may reduce the profits that many community banks earn on small business loans.
However, some community banks should be able to exploit profitable opportunities presented by credit scoring. For example, a bank could focus its efforts on lending to those small business borrowers who want the personal contact that the credit-scoring system does not provide. In addition, some small business borrowers will not wantor may not qualify forthe standard loan that everyone else is getting. Community banks can try to cater to these nonconforming borrowers who will likely have to pay more for their nonstandard borrowings. Finally, a community bank can try to increase its lending via a credit-scoring system to make up for the lower returns it earns per loan. However, the relatively limited number of businesses that community banks serve and their inability to match the cost efficiencies and investment capabilities of larger banks going after the same markets limits the opportunity presented by these possibilities.
Predicting the future of financial services is a low-success venture. Given that caveat, it does seem more likely than not that credit scoring will significantly change the borrowing opportunities for small businesses. Moreover, credit scoring small business loans seems to pose a distinct threat to the profitability of some community banks. The greatest unknown, however, is how the changes wrought by credit scoring will alter the nature of community commerce in rural areas and small towns. Will community life be modified significantly by the severing of the personal relationship between local lenders and local small businesses? Will the benefits that accrue to some borrowers who have greater access to cheaper funds offset the costs to community life and local banks?
There are no obvious answers to such questions. In fact, it would seem unlikely that a technical change like credit scoring could produce serious changes to community commerce. Yet it is clear that changes in technology can alter the processing of information and the distribution of products and that such changes have already affected commerce in rural areas and small towns. One need not look further than the effect of the Wal-Mart chain of discount stores, with its sophisticated distribution and information management systems, on small communities. Credit scoring is probably best understood as another of the technological advances that will provide benefitsand coststo nonmetropolitan communities.
Ron Feldman is a senior financial specialist in the Special Studies Unit of the Minneapolis Reserve Bank's Banking Supervision Department. He studies trends affecting financial institutions and analyzes banking and supervision policies.