By Jay Vadiveloo
This is the third of a series of articles in Contingencies on small businesses in the U.S., abbreviated as SMEs, or small and medium-sized enterprises. The first article, “Putting It to Work—Risk management services for vulnerable small businesses,” introduced the Students for Workers Movement (SWM) initiative by the Goldenson Center, in which teams of actuarial students—at no charge to the SME—did a risk management analysis of vulnerable SMEs in the region, culminating in a report on ways to mitigate risks and take advantage of revenue-generating opportunities. The SWM initiative focused on vulnerable SMEs including minority-owned businesses, women-owned businesses, SMEs owned by veterans and SMEs with disabled owners.
The success and publicity we received from the SWM article led to the second article, which was published in the March/April 2024 issue, “Small Business Life Expectancy—Measuring the sustainability of a small business.” Using recent data from the Bureau of Labor Statistics (BLS)—which provides annual survival rates of SMEs for each calendar year, by each state, and by SME category—a single risk metric, the Small Business Life Expectancy (or SBLE), was calculated. The SBLE calculations were done and ranked for each state and for each small business category. We’ve seen, as a consequence of that article, a lot of interest in identifying states that have the most conducive environment for the sustainability of small businesses and the small business category that has the longest life expectancy.
In this article, we have developed an algorithm to calculate the SBLE for any small business that captures the individual characteristics of how the small business is run and managed. For example, the SBLE for a small business category in a given state may be six years, but for an individual small business in that same business category in the state, the customized SBLE could be 7.5 years based on some positive measures used by the small-business owner to manage the business.
Our Model
To customize the SBLE calculation for any small business, we used a combination quantitative and qualitative approach to develop the customized SBLE. The quantitative component of our model used the BLS database and the methodology in our second article to develop the baseline SBLE by state and business category. The qualitative component of our model utilized a Delphi-type approach to research the top 10 factors that result in failure of a small business and the corresponding weights for each factor.
We used, as a reference for this qualitative piece of our model, the research report “The Top 12 Reasons Startups Fail” by CB Insights, available online here. Some of the top factors included issues like lack of capital, not meeting market needs, competition, regulatory issues, etc. We used the small-business owner’s response to these qualitative factors, together with the factor weights, to adjust the baseline annual failure rates to develop a customized SBLE for the specific small business.
We determined the importance of each of these 10 factors by having the business owner rank them on a scale of 1 to 5, with 1 indicating that the factor was a significant uncontrolled risk for the small business and 5 indicating that the risk factor was well under control. Based on the response for each risk factor and the corresponding weights, for each factor, an adjustment was made to the baseline rate—and then all these adjustments were multiplied together to come up with an overall adjustment factor. To ensure the overall adjustment factor was reasonable, we capped the adjustment for the overall risk factor to be within 1 standard deviation of the baseline SBLE.
SBLE Customization Illustration
Consider a small startup restaurant business in Connecticut. Based on the BLS, the estimated baseline life expectancy is 10.3 years using the annual baseline failure rates by state and by business category from the BLS. All the 10 risk factors were ranked on a scale of 1 to 5, where 1 represents the greatest risk level and 5 represents the smallest risk level. The annual baseline failure rates are then adjusted as follows:
While these adjustment factors are subjective, they are consistent and logical in that annual failure rates are reduced (or increased) in proportion to the owner’s assessment of the significance of the risk. No adjustment is made to the annual baseline failure rates for a ranking of 3. The aggregate response risk adjustment to the annual baseline failure rates is simply the product of the response adjustment rates for each of the 10 risk factors.
The reference article provides weights for each of the 10 risk factors. The weight for each of the 10 risk factors is normalized to sum to 1 and a simple but logical rule is adopted to reflect the weight (i.e., importance) of each risk factor in determining the final risk adjusted rate. The rule is determined so that a response of 5 for an “important” factor will have a greater risk-adjusted annual survival rate than the same response for a less-important factor. In contrast, a response of 1 for an “important” factor will have a smaller risk-adjusted annual survival rate than the same response for a less-important factor. The rule can be described as follows:
For a given risk factor with normalized weight w (w<1), for response 4 and 5, the risk adjusted survival rate is calculated as:
(annual baseline survival rate)^[(response risk adjustment)*(1-weight)]
For response 1 and 2, the risk adjusted rate is calculated as:
(annual baseline survival rate)^[(response risk adjustment)*(1+weight)]
For response 3, there is no adjustment to the annual baseline failure rate.
The customized SBLE for the specific restaurant in Connecticut is determined by using the annual risk-adjusted survival rates to determine the SBLE for the small business. The corresponding customized SBLE is further adjusted to ensure that it lies within 1 standard deviation of the baseline SBLE.
Example
Consider the same small restaurant business in Connecticut and assume the responses to the 10 risk factors, baseline failure rates, and weights are as follows:
Greater adjustments mean worse survival probability.
Smaller adjustments mean better survival probability.
Aggregate risk-adjustment factor = 132.67%
Baseline SBLE = 10.3 years
Customized SBLE = 7.8 years
Sensitivity Analysis
Consider the same example of the startup small restaurant business in Connecticut. The baseline SBLE from the BLS is estimated as 10.3 years.
Assume the response is 5 for each of the 10 risk factors—i.e., these risks are all well within the control of the small-business owner. Then the customized SBLE increases to more than 20.9 years. That’s an increase of at least 10.6 years over the baseline SBLE.
Now assume the response is 1 for each of the 10 risk factors—i.e., all the risks to the small business are outside the control of the small-business owner. Then the customized SBLE decreases to less than 1 year. That’s a reduction of more than 9.6 years from the baseline SBLE.
So depending on the business owner’s response to these qualitative risk factors, the customized SBLE could range from 9.6 years lower or 10.6 years higher than the baseline SBLE.
Conclusions
The customized SBLE model and the app we have developed is a new metric that any small-business owner in any state or in any business category could utilize and benefit from its findings. Besides providing a single number to evaluate the sustainability of the small business, it could also help the small-business owner to determine which risk factors to focus on in order to increase the sustainability of their small business. While some of the adjustment factors are subjective, it is logical and consistent and correctly captures the riskiness and sustainability of the small business and provides valuable insights as to what corrective measures are needed to reduce risk and increase the SBLE metric.
There is a famous quote about risk-taking that states that the biggest risk is not taking any risk. Our model is not designed to stifle risk-taking by small business owners; rather, it provides them with a tool to better measure and manage the risks small businesses are exposed to.
Acknowledgments
The author is grateful to all the students who worked on this model and helped to consolidate the Bureau of Labor Statistics data—Jizhou (Nelson) Chen, Shuyao (Zoe) Cai, Chung (Catherine) Yun, Abigail Baidoo, and Zhiguo Wang, my assistant director. We hope that this article and the app that we have developed will help the small business community better understand, measure, and manage the myriad risks small businesses are exposed to.
JAY VADIVELOO, MAAA, FSA, is the director of the Janet & Mark L. Goldenson Center for Actuarial Research at the University of Connecticut.
