Skip to content
MarginGraph

The private company discount, and the case against it

Being unlisted is supposed to cost you 20 to 30 percent of your value. The published estimates run from zero to fifty, and the best-identified ones cluster far below what practitioners apply.

6 min readMarginGraph

There is a number that gets applied to almost every private business valuation, usually without argument. Your business is not listed on a stock exchange, so a discount comes off. Twenty percent, thirty, sometimes more. The reasoning is that shares you can sell in a click are worth more than a business that takes a year to sell.

The size of that discount is not a detail. On a business valued at €2m before it is applied, the difference between 7% and 30% is the difference between losing €140,000 and losing €600,000. Both of those percentages come from published studies of real transactions.

So the question is which study, and what it actually measured. The literature here is genuine, and it is considerably less settled than the number suggests.

The case for

The most-cited transaction study is Micah Officer's, published in the Journal of Financial Economics in 2007. It documents "acquisition discounts for stand-alone private firms and subsidiaries of other firms (unlisted targets) of 15 to 30%, on average, relative to acquisition multiples for comparable publicly-traded targets," and attributes the gap primarily to liquidity rather than to information asymmetry.

The practitioner mainstay is different: restricted stock studies. When a listed company places shares privately, those shares cannot be resold for a period, and they change hands below the market price. The gap is treated as the price of illiquidity.

Stout Restricted Stock Study
15.7%
median discount, 783 transactions, 1980 to 2025
Same study, mean
20.4%
the mean sits well above the median
FMV study, Robak & Hall
20.1%
median across 230 transactions, 1980 to 1997

Those are real numbers from real transactions, and they are the basis on which roughly three-quarters of valuation practitioners apply a discount.

The case against

The population is wrong. Every restricted stock study measures shares of an already public company that cannot be sold for a defined period. That is a temporary registration restriction with a known end date. Your business is permanently unlisted, has no quoted price, and cannot be sold in a click at any point. Whether one is a good proxy for the other is an assumption, not a finding.

The selection is not random. Aswath Damodaran's review of the standard studies is blunt on this. The samples are small and spread over long periods with substantial standard errors. Firms that do private placements "tend to be smaller, riskier and less healthy than the typical firm," so the discount picks up distress as well as illiquidity. And investors buying into a private placement often provide other services to the company, for which the discount may be partial compensation.

Controlled studies find much less. The older studies compare a restricted share price with a market price and call the whole gap illiquidity. Once you hold the other differences constant, the part attributable to marketability itself shrinks sharply: one such study puts it at 7.23%, another at a 10.4% median. Against the 30% to 35% the uncontrolled studies report, that is a different claim about a different thing.

Pre-IPO studies should not be used at all. These compare private transactions in a company's shares with its later IPO price, and they produce the largest discounts: Emory's studies report 40% to 50%, and one Willamette series shows a median around 50%. Gilbert Matthews, a valuation practitioner with 35 years at Bear Stearns, lists six fatal defects, including that the IPO price is unknowable at the time of the earlier transaction, that IPO underpricing inflates the apparent discount by an average first-day gain of over 15%, and that companies which never went public are excluded entirely. In 2001 the same Willamette series produced a median of minus 196%. Roughly 40% of practitioners still use these studies.

And the strongest paper finds nothing. In 2019 a team of four finance academics redid the transaction work in the Journal of Financial and Quantitative Analysis, correcting two statistical problems in the earlier studies. Their abstract is worth reading twice:

Academic literature, practitioners, courts, and regulators routinely assert that both private and subsidiary targets sell at discounts relative to public targets. However, the empirical evidence to support this conclusion is thin. Our work alters the methodology from prior research to avoid biases due to both one-sided sample truncation and Jensen's inequality. Following these changes, we find no evidence that unlisted targets sell at discounts.

Jaffe and colleagues, Journal of Financial and Quantitative Analysis, 2019

Two specific statistical corrections, and the effect disappears.

Where that leaves the number

MethodReported discountWhat it actually measured
Pre-IPO studies40% – 50%Rejected by senior practitioners on six grounds
Uncontrolled restricted stock30% – 35%Listed shares under temporary sale restriction
Recent restricted stock16% – 20%Same population, larger and cleaner sample
Transaction comparison15% – 30%Unlisted targets versus comparable listed targets
Controlled for firm traits7% – 10%Marketability isolated from distress and size
Bias-corrected0%No discount detectable

The estimates that survive methodological scrutiny cluster between 7% and 20%. The large numbers come from studies whose authors' critics, and in some cases whose own summarisers, acknowledge selection bias.

What to do

Insist on a named source. Any discount applied to your business should come with a study, a population and a year. "Standard practice" is not a citation.

Check the direction of the bias. If the source is a restricted stock study, its firms were smaller, riskier and less healthy than average. That is a reason the number is high, not a reason your business deserves it.

Argue about the mechanism instead. Illiquidity has a specific meaning: how long it takes to find a buyer and how much price you give up for speed. Those are facts about your market, your sector and your size, and they are arguable with evidence. A fixed percentage is not.

Do not accept it twice. If your comparables are already private company transactions, and most published SME multiples are, the discount is already inside them. Applying a marketability discount on top of a private company multiple double-counts, and it is a common enough error to be worth checking for explicitly.

What this does not tell you

None of these studies is about a business like yours. Officer's compares acquisition multiples of unlisted targets with listed ones, mostly at a scale far above an owner-managed business. The restricted stock studies are about listed equity. The bias-corrected study that found nothing was working in the same large-deal universe.

There is no study of what an owner-managed business with €500,000 of earnings sells for relative to some liquid equivalent, because no liquid equivalent exists. Everything above is an argument by analogy, and knowing that is more useful than any of the percentages.

Every adjustment, stated and sourced

The report shows which adjustments were applied to your figures, why, and what the range looks like without them.

Sources

A note on a source we did not quote a number from

Koeplin, Sarin and Shapiro's 2000 paper in the Journal of Applied Corporate Finance is the foundational transaction-based study on this topic and is cited everywhere, usually with a figure attached. We could not confirm those percentages from any freely available source, so we have left the number out rather than repeat what everyone else repeats.

Frequently asked

Decision2 min

What is my business worth?

Upload your financials and receive a valuation report with assumptions, risks and a valuation range. Three methods, every figure traced back to a line in your file.

9Generate Report

Article7 min

The number in the middle that nobody can source

Every risk that supposedly lowers your business value passes through one figure in the discount rate. The profession's own guidance says there is no data source for it.

Article6 min

Where SME multiples actually come from

Every multiple you have read was published by someone selling something. Here is who collects the data, how, and why two datasets covering the same businesses disagree by a factor of two.