Franchises Have Higher SBA Loan Failure Rates
But Independent Experts Say Study Misleading
WASHINGTON, D.C. (Blue MauMau) - A study concludes that SBA loans given to franchises had a higher rate of repayment failure than loans to non franchises (see chart). Franchises had a 6.5% charge-off rate, the amount that the lender has to dismiss or charge off for having a nonpaying bad loan, which was more than the 5.9% for the average small business.
Chart provided by permission of the International
Franchise Association's journal, Franchising World
The International Franchise Association's Education Foundation contracted with Frandata Corp., a franchise industry research firm, to compare Small Business Administration (SBA) non-franchise loans to franchise loans from 2001 through 2006. Lenders can receive a guaranty of payment by the SBA if they structure their loans according to the Administration's requirements.
Frandata also concluded in the study that the sectors with the highest loan failures or charge-off rates were in the gift & novelty, sporting goods, automobile and hobby & toy retail segments.
Some 406,000 SBA loan disbursals to businesses were looked at for the six-year period, of which 25,744 were to franchise businesses. Regarding these disbursals, the research firm declares, "Based on the expanded data, we were able to compare, by industry, franchise loan loss compared to SBA total loan loss . . . . Franchise charge-off percentages were higher than SBA total charge-off percentages in four years and were lower in two years."
The charge-off rate is the dollars charged off divided by the total dollars disbursed.
Mr. Darrell Johnson, President of Frandata, concludes, "There is not much difference in SBA loan guarantee failure rates between franchise and non-franchised businesses. In lay terms, slightly less than 6 of every 100 SBA loans for non-franchised businesses go into a work-out with the bank/SBA, compared to about 6 ½ of every 100 SBA loans for franchised businesses."
Said one franchise insider, "Showing franchise units that fail at paying back their SBA loans at roughly the same rate as non-franchised units in itself is a major shift for the International Franchise Association, who for years declared that buying a franchise was a proven model and much safer than investing in a non-franchised business."
Problems With Study
But there is a defect with the reliability of the SBA numbers; namely, Frandata says that the SBA may have lenders who do not properly identify borrowers as franchises.
According to some experts, that is a key flaw of this study.
Frandata notes in the report, "By comparing specific franchise system unit levels to SBA data for the same systems, Frandata determined that there may be significant underreporting of SBA lending which could affect any comparison of franchise loan performance to non-franchise loan performance. The SBA determines if a loan is a franchise loan when the business applying for an SBA loan or the lender identifies the borrower as a franchise in the loan documentation. Because some businesses and some lenders do not properly identify borrowers as franchises, the number of franchise loans may be underreported. Frandata is working with the SBA to address the underreporting issue."
It is uncertain what results a cleaning up of the data would yield.
If the SBA loans to franchisees are underreported, as Frandata suspects, it could mean that there may actually be more franchises who fail to pay back their SBA franchise loans than non-franchises. It is also plausible that the charge-offs for franchises could be considerably less than independents. Or the new results could come out as fairly equal again.
When asked about how any study that made conclusions could be released if the franchise loans were suspected of being underreported, Mr. Johnson stated, "Since the data and the accuracy comments came from the SBA, I’d defer to them on this question."
Franchise and statistics experts chime in that it is difficult to assess the reliability of results based on imperfect data. Statisticians spend considerable time cleaning up data, particularly when there is a misalignment of groups. It is difficult to give any sort of confidence level, a statistical tool that tells you how sure of the numbers to be, without dealing with this issue first.
This study comes with no confidence level at all.
Scott Shane, Professor of Entrepreneurial Studies at Case Western Reserve University, who is well-known for statistical studies in franchising, notes another problem. Even if the numbers were good, he explains why the Frandata conclusion is misleading.
States Professor Shane, "Let me just give one example of why. What if the reason that franchised rates have higher charge-off rates is because they are, on average, physically bigger restaurants. The bigger loans they carry might be the causal factor behind the charge-off rate. It's possible that the study controlled for these differences — again I didn't see the study itself so I don't know — but even if it did, I could start identifying a host of differences that they didn't control for. "
"Suppose the charge-off rate is higher for franchised restaurants than non-franchised ones, which at least is consistent with the aggregate data. If, and this is a really big if, franchised restaurants are the same as non-franchised ones on all other dimensions, then you could interpret this. But we know that's not true. The types of food served, size, geographic location and a host of other things are actually different for franchised and non-franchised restaurants."
--
Related readings:
- A Study of Franchise Loan Performance in the SBA Loan Guaranty Programs by Darrell Johnson and John Reynolds, Franchising World
(Reporter's note: See Key Conclusions and Methodology at the bottom of the article) - Charts Showing Conclusions of the Study (one page pdf file)
Study is misleading
For a study like this to be really valuable, to anyone trying to decide between franchise and non-franchise, the data should be segmented by length of time in business as a franchise system. As an example, what was the failure rate for franchise systems in business 10 years with 500 units? How did that failure rate compare with a system 2 years in business and 20 locations.
Furthermore, if 5 locations in a 10 unit franchise system fail, one could not assume that the franchise system was "proven".
Bob Rodi, CLP
President
Mount Pleasant Capital Corp.
www.mountpleasantcapital.com
Benefits of self-delusion
The flaws noted in the article and by Mr. Rodi are valid points. To those, I would also add that even if the data are correct, this does not necessarily mean that higher rates of franchisee failure are due to anything related to the system itself. It might be related to the type of person who purchases a franchise versus the person who goes independent.
Many people buy a franchise because of risk-aversion, believing that they are safer and that they are entering into a "partnership" with the franchisor. Many franchisors do infantilize their zees, but let's just say that the zees are often willing to be infantilized. Now, I would dispute that franchising is inherently safer than going independent, but that is the perception, and perception causes a personality-type bias in those who select franchising versus going independent.
Conversely, someone going it alone is likely to be more comfortable with risk, and more confident--even over-confident--in their abilities. In contrast to a franchisee, the independent may retain a belief that his future is in his hands, and that there is no one to blame for failure except himself.
Such personality differences can affect the likelihood of success in business. The very illusions which cause the independent to go out on the wire without a net may result in a difference in success rate.
Many of my successful business owners (franchised and non-franchised) have a healthy dose of over-confidence. It is worth considering this comment by a scientist who studies "rational actor" behavior: