Owner dependence: strong logic, no price data

Every buyer asks some version of the same question: what happens to this business if you are not in it? Nobody can price the answer, which does not stop people trying. Ask what owner dependence costs and you will be told a full turn on the multiple, or ten to twenty-five percent, or that it removes half the buyer pool.

We could not find a published estimate behind any of those. What we did find answers the question the discount was only ever standing in for: does the person at the top actually move the numbers?

#The natural experiment

The problem with measuring whether a boss matters is obvious: good businesses attract good leaders, so finding the two together tells you nothing about which caused which. In 2020 three economists published a study in the Journal of Finance that got around it in an unpleasant but rigorous way. They used CEO hospitalisations: an absence nobody chose, which lets you watch the same business with and without its leader.

The sample is unusually well matched to the readers of this article: roughly 13,000 Danish small and medium-sized businesses, observed from 1996 to 2012.

Longer absences

≈1 pp

decline in operating return on assets at 10+ days

Shorter absences

measurable

even 5 to 7 days moved operating performance

Other executives

no effect

the pattern did not repeat for non-CEO absences

Operating return on assets is trading profit measured against the assets used to earn it. A one point fall is not a catastrophe on its own. What matters is that it moved at all, that it moved in the direction everyone assumes but nobody had shown, and that it only moved for the person at the top.

Three findings from the paper matter for an owner thinking about selling.

The effect is real and it is specific to the top job. In the authors' words, "CEOs are unique: the hospitalization of other senior executives does not have similar effects." That is the empirical version of the thing buyers worry about.

It is worse in exactly the businesses most likely to be sold by their founder. The effects are "larger for younger CEOs, in growing and family-controlled firms, and in human-capital-intensive industries." A growing, owner-run service business is the profile with the most to lose.

The authors' conclusion is a management instruction, not a discount. "CEO contingency plans are valuable."

#What is missing

That study measures operating performance while the CEO is away. It does not measure what a business sells for.

There is no published estimate of a key person discount on a small business transaction price that we could locate. The concept appears in the practitioner literature, notably in Pratt's work on discounts and premiums, as a synthesis of United States tax court outcomes. That is a record of what courts accepted in specific disputes, not an empirical estimate of market behaviour.

#Why the discount was the wrong question anyway

A discount is a way of pricing something you cannot fix during the transaction. Owner dependence is not that kind of risk. It is the one significant valuation factor that is almost entirely within your control, and the only real constraint is time.

That reframing is worth more than a number. If owner dependence cost a fixed percentage, the rational response would be to accept it and negotiate. Because it is fixable, the rational response is to start early, and "early" here means at least a year.

#What actually reduces it

The evidence above points at what a buyer is really pricing, which is not your presence but the consequence of your absence. Four things change that consequence, roughly in order of how long they take.

• Document the decisions only you make, and the basis on which you make them. Not the tasks, the judgement calls.
• Move customer relationships to at least one other named person, with the customer's knowledge.
• Put a second signature on anything that only you can currently authorise.
• Take a two-week absence and let the business run without contact. That is the test, and it produces evidence a buyer can verify.

The last one is the closest thing to proof you can offer. A buyer's concern is unfalsifiable while you are in the building every day. An owner who can point at a fortnight where nothing broke has converted an argument into a fact.

#What a buyer will do with your answer

Expect structure rather than price. When a risk cannot be underwritten, buyers do not usually discount it, they shift it. That means an earn-out, a longer handover period, a consultancy agreement, or part of the price held back against your continued involvement.

That is where the money actually shows up, and it is worth seeing the size of it. Nobody publishes an average, so treat the following as arithmetic rather than evidence: on a €1.2m price, a buyer holding back a fifth against your staying for two more years is €240,000 you do not bank at completion, and might not bank at all. Compare that with any percentage discount you have read about, and you can see why the structure is the conversation.

All of it says the same thing: the buyer will not pay today for performance that depends on you being there tomorrow. Reducing dependence before the process starts is what turns that into a price conversation instead.

#What this does not tell you

The Danish study measures operating return on assets during absence. It does not tell you what your business would sell for with or without you, and no study does.

It also comes from a population of businesses that had a CEO to hospitalise, which excludes the smallest owner-operated firms where the owner is the entire operation. If that is your business, the effect is presumably larger, and there is no evidence at all about it.

And there is an uncomfortable implication for anyone selling a business that genuinely depends on them. If the owner is the product, then what is being sold is a job with equipment attached. That is a real thing to sell, and it is priced as a job, which is part of why smaller businesses transact on seller's discretionary earnings rather than EBITDA.

#Sources

• Morten Bennedsen, Francisco Pérez-González and Daniel Wolfenzon, "Do CEOs Matter? Evidence from Hospitalization Events," Journal of Finance 75(4), 2020, pp. 1877–1911. https://econpapers.repec.org/RePEc:bla:jfinan:v:75:y:2020:i:4:p:1877-1911
• INSEAD, summary of the same research, for the sample of approximately 13,000 Danish businesses between 1996 and 2012 and the magnitude of the operating return decline. https://www.insead.edu/news/insead-research-finds-how-much-ceos-matter-firm-performance
• Shannon Pratt, Business Valuation Discounts and Premiums, Wiley, chapter 17, for the practitioner treatment of key person discounts as a synthesis of United States tax court outcomes rather than an empirical estimate.

#A note on what we left out

Key person discount percentages in the range of ten to twenty-five percent are widely quoted. We found no empirical study estimating a key person discount on small business transaction prices, so we have not repeated a figure. The evidence in this article is about operating performance during a CEO's absence, which is a different measurement, and we have said so rather than letting one stand in for the other.