When Smaller is Better

The Effect of Bank Size on Small-Business Lending

Research by Raghuram Rajan

According to new research, small banks have a comparative advantage in the arena of small-business lending because they are better able to collect and act on so-called "soft information" than large banks.

Due to changes in technology and the ongoing consolidation of the commercial banking industry over the past thirty years, the relationship between banks and borrowers has been growing more distant and impersonal. These changes raise questions not only about whether large banks will behave differently than the smaller banks they are displacing, but about the differences between large and small organizations as a whole.

One reason to believe that large organizations may behave differently than small organizations is that they may have different abilities to process hard and soft information. Hard information, such as audited earnings, is easily captured on paper. Soft information-intangible factors such as a potential client's strength of character-is difficult to communicate.

The study "Does Function Follow Organizational Form? Evidence From the Lending Practices of Large and Small Banks," by Raghuram Rajan, a professor at the University of Chicago Graduate School of Business, Allen N. Berger and Nathan H. Miller of the Board of Governors of the Federal Reserve System, Mitchell A. Peterson of Northwestern University, and Jeremy C. Stein of Harvard University examines whether large banks do as well as small banks in small-business lending, an activity that relies heavily on collecting soft information about borrowers.

"We wanted to see whether bank size has any effect on the nature of small-business lending," says Rajan. "And what we found was that bank size has a tremendous effect on the nature of the relationship between a small firm and its bank."

Hard and Soft Information

The reason for this effect is that the ease with which hard and soft information can be used is often determined by the size of the bank. Information about a small-business client has to travel a much greater distance in a large banking conglomerate-from the loan officer in a branch office to the decision-making authority in a far away city-than in a small town bank. Only hard information, in the form of easily verifiable data such as income statements, balance sheets, and credit ratings, can be credibly passed along such a bank hierarchy.

However, 43 percent of the firms in the study do not have formal financial records, indicating that soft information plays a significant role in determining whether these firms are creditworthy. In many cases, these firms represent "main street America," such as the local barber who needs financing to set up an additional chair in his shop. Record keeping at many of these establishments may be haphazard at best.

When loan officers have very little hard information to work with, they have to estimate a client's earnings and judge intangible elements such as trustworthiness to determine if the potential client is a candidate for a "character loan." These evaluations may be derived from subjective qualities such as the strength of the client's handshake or how well they make eye contact.

"The face-to-face interaction may help the loan officer form a picture about whether the client can be trusted," says Rajan, "but when it comes time to write a report, they're only writing down a sentence for their bosses saying if this person is trustworthy or not."

At small banks, the decision-making authority is likely to be in close proximity to the point of information collection, which facilitates the communication of soft information. For example, when a client requests a loan from a small bank, the loan officer can easily bring the client to meet with his manager face to face. The lending decision can then be informed by the soft information resulting from this meeting.

However, if the loan officer works in a small town branch office of a large bank headquartered thousands of miles away, it is more difficult for that loan officer to communicate soft information in a way that captures the richness of that information.

Serving Different Niches

To study the relationship between bank size and small-business lending, Rajan and his coauthors used the Federal Reserve's 1993 National Survey of Small Business Finance, which examined the financing practices of a wide sample of small firms. The sample contains 1,131 for-profit firms, all of which have fewer than 500 employees.

The authors looked at the firm's most recent bank loan and matched each firm with the specific bank from which it borrowed. Information about the banks was obtained from the Consolidated Report of Condition and Income and the FDIC Summary of Deposits.

The authors' findings include the following: First, bigger banks are more apt to lend to firms that are larger or that have more detailed accounting records. Second, the physical distance between a firm and the branch office that it deals with increases with the size of the bank. This is consistent with notion that large banks rely less on the soft information that is typically available through personal contact and observation. Third, firms do business with large banks in more impersonal ways, communicating more by mail and telephone than face-to-face meetings. Fourth, bank-firm relationships tend to be both longer-lived and more exclusive when the firm in question borrows from a small bank. The reason for this last finding appears to be that soft information produced by small banks is more likely to be specific to a given banker and borrower, and not easily transferable. Accumulated soft information binds a borrower to its bank over time.

Given that small banks are better at building relationships based on soft information, it can be expected that firms that are bigger credit risks (including "difficult" borrowers, such as firms that do not keep formal financial records) or require more nurturing will seek loans from small banks, ensuring that small banks occupy a niche market that will survive the onslaught of bank mergers.

While the size of the bank that a firm borrows from matters, the authors find that the size of the bank's holding company does not. Lending decisions are typically made at the bank level rather than higher up at the bank holding company level. This pattern suggests that it is not simply the absolute size of an organization that is important, but also the degree of decentralization that can be achieved. "Large organizations can undertake activities involving soft information if they decentralize," says Rajan.

Implications for Other Industries

The study's findings regarding organizational size and the use of information apply to other industries that emphasize relationship building, such as consulting, law, research and new product development, and law enforcement. In all these industries, the organizational structure may play a crucial role in determining how effectively the job is carried out.

The study also suggests that if multinational banks substantially crowd out a developing country's domestic banks, the supply of loans to small firms could be reduced. In these cases, Rajan recommends that the headquarters of the foreign bank should allow the local branch to have a reasonable amount of decision-making authority over the loans.

In addition, Rajan and his coauthors caution that in many countries, the standard practice of setting up large bureaucratic organizations to provide subsidized credit to small businesses may not be very effective. "In a situation where soft information is generated, bureaucracy does stand in the way, because bureaucracies are incapable of processing the information that small-business lending is based on," says Rajan.

Preliminary evidence by these researchers suggests that the credible decentralization of decision making can offset the negative effects of large organizations' size. Therefore, it may be possible for a large organization to enjoy the best of both worlds if it sets up internal structures that achieve the right level of decentralization.

Raghuram Rajan is Joseph L. Gidwitz Professor of Finance at the University of Chicago Graduate School of Business.

>> Request a copy of the research paper featured in this article