SBA Loan Performance and the Challenges that lie in finding the most accurate information

January 18th, 2019 by Ritwik Donde

W. Edwards Deming, whose sampling techniques are still used by the U.S. Department of the Census and the Bureau of Labor Statistics, once said, “In God we trust; all others must bring data.” What he did not know then was what happens if it is “bad data.”

The last time the U.S. Small Business Administration (SBA) officially released a consolidated list of loan-loss performance for franchise brands was 2011. For years franchise lenders have used this SBA failure rate data as an initial screening tool to determine whether they should consider a loan to a prospective or existing franchisee.

Since 2011, the raw data about SBA loan performance has been published on a quarterly basis. However, identification of franchise loans in this data set comes with its own challenges. Starting in fiscal year 2018, the SBA has started tagging the franchise loans in three ways:

  • old identifier codes which were assigned prior to the incorporation of an organized numbering system;
  • FRUNS codes (Franchise Unique Number System as incorporated by FRANdata) and,
  • New codes assigned as per the SBA Directory.

This makes it very difficult for underwriters and franchisors to accurately identify their loan performance statistics. There are also other challenges with SBA franchise such as: a) loans that should have been attributed to franchise brands have not been; b) loans that have been attributed to franchise brands but should not have been; c) loans that have been attributed to the wrong franchise brands; and d) loans that have been attributed with a unique identifier code called “TEMPORARY FRANCHISES” leaving gaps in connecting these loans with the appropriate franchise brand.

For example, when we examined the most recent data for SBA loan performance, we found an instance where a loan was coded under “93137”. This identifier code does not match with the new SBA Code on the SBA Directory or FRUNS, making it impossible to accurately match the underlying franchise associated with this loan. In another case, we found that the SBA Directory code was wrong entered as “S2075” for a brand’s loan, which resulted in under-reported loan performance.

A recent FRANdata examination of SBA data discovered more than 400 loans made to franchisees of a major sub-franchise which were not labeled as franchise loans and were not grouped under the SBA Code for the brand. Our review of the list of brands that the SBA identifies as franchises found more than 500 which we have not been able to verify as franchises, even considering the SBA’s broad definition of what constitutes a franchise that expands beyond business format franchises. In working with separate brands with similar names, we found numerous loans attributed to the wrong brand.

In light of such challenges, lenders are increasingly relying on FRANdata’s FUND Scores to more accurately assess a brand’s performance. This is mainly because unit outcomes are a more complete picture of how a brand is performing, as SBA loan performance information not capturing all the performing loans made to a particular brand distorts brand performance.

Categories