Why Small Private Companies Sell for Lower Multiples

Why the multiple depends on which earnings you divide by

A multiple is a fraction. The price on top is usually clear; the earnings figure underneath is where most published ranges stop being comparable. Before asking why small companies sell for less, check what each dataset divided by.

The IBBA and M&A Source Market Pulse survey states its basis in the chart footnote: deals from under $500K to $2M in purchase price are "reflected as multiple of SDE (Seller's Discretionary Earnings)," and deals from $2M to $50M "as multiple of EBITDA" ^[1]. The switch happens inside the same chart, at the same $2M boundary where the survey moves from "Main Street" to "Lower Middle Market."

The two measures answer different questions:

• SDE, as brokers commonly use it, is the cash a single owner-operator takes out of the business: pre-tax earnings with the owner's own pay and discretionary expenses added back. It suits a buyer who will run the business personally.
• EBITDA treats management as a cost. A buyer who has to hire a general manager divides the price by earnings after that salary.
• Adjusted EBITDA adds back items the seller argues are non-recurring. Its size depends on who prepared the adjustments and whether a quality-of-earnings review tested them.

For the same company, SDE is larger than EBITDA whenever the owner draws a salary. A lower multiple of a larger number can therefore describe the same price as a higher multiple of a smaller one. Part of what looks like a "small company discount" in a combined chart is simply this change of denominator.

What the largest public datasets say, and who is in them

Two broker-sourced series are widely quoted. Neither is a census of transactions, and they should not be spliced together.

BizBuySell Insight Report. The current BizBuySell Insight Report page carries a "Q3" report; its methodology note says year-over-year comparisons use revised Q3 2025 figures, and the page's archive lists the 2026 first- and second-quarter reports as past editions, so the period is Q3 2026 (checked 2026-10-28). The report splits sold businesses at a $1 million sale price:

Source: BizBuySell Insight Report, accessed 2026-10-28.

The same report says businesses selling above $1 million made up 22% of Q3 transactions, "the highest share on record," and that the market-wide median sale price "rose 17% year-over-year to $350,000," partly because of "a shift in deal mix rather than rising prices across the board" ^[2]. That sentence is a warning that applies to every average in the table: the mix moves the number.

Who is in it: the report is based on "approximately 50,000 businesses for sale and those recently sold," and closed transactions "are reported to BizBuySell.com on a voluntary basis by business brokers nationwide" ^[2]. Deals that close without a broker, or with a broker who does not report, are absent. The page does not define "cash flow." One broker quoted in the report refers to "SDE multiples" for Main Street businesses, which suggests the series sits closer to SDE than to EBITDA, but the page does not establish it.

IBBA and M&A Source Market Pulse. The Q2 2026 edition, the most recent we found on 2026-10-28, "was conducted July 1-15, 2026 and was completed by 255 Business Brokers and M&A Advisors," who "completed 181 transactions this quarter" ^[3]. The release says the $5 million to $50 million segment "climbed from 5.5 to 5.8 (its highest since Q1 2022)." The Highlights chart gives Q2 values for each segment:

Source: IBBA, Market Pulse Q2 2026 Highlights, "Business Value Multiples Q2 2023 - 2026" chart.

The Highlights page does not say whether these are medians or averages. Its sample is 181 deals across five segments, so each cell rests on a small and unstated count. The ladder still rises with every step in size, and the step at $2M is on a different earnings basis from the steps below it.

Read together, the two series agree on direction and disagree on almost everything else: different sources, different size cut-offs (sale price at $1M versus purchase-price bands), a "cash flow" measure on one side and an SDE/EBITDA split on the other, a quarter apart in time. They are two non-comparable series that point the same way, not one curve.

The private company discount literature

The academic version of the question is narrower: do private companies sell for less than otherwise similar public companies? John Koeplin, Atulya Sarin and Alan C. Shapiro addressed it directly in "The Private Company Discount," Journal of Applied Corporate Finance 12, no. 4 (2000): 94-101. Its subject is exactly the gap this article is about: private targets priced against public comparables.

The paper compares four valuation multiples for a set of private-company acquisitions and a comparable set of public-company acquisitions. Its abstract, on the publisher's page, reports that "Domestic private companies are acquired at an average 20–30% discount relative to similar public companies when using earnings (more precisely, EBIT and EBITDA) multiples," that the discount on price-to-book multiples is "somewhat lower," and that there are "no significant differences between the revenue multiples of acquired private and public companies." For non-U.S. targets the discount on earnings multiples was 40-50%, though less statistically significant ^[4]. The authors attribute part of the gap to "fundamental" factors that kept the firm private, not only illiquidity. These are pre-2000 averages comparing private with public targets, not small private companies with larger ones.

Practice has since hardened findings like these into a single number. Aswath Damodaran's NYU Stern slides on private company valuation describe the field's habit plainly: "All the talk, though, seems to lead to a rule of thumb. The illiquidity discount for a private firm is between 20-30% and does not vary across private firms" ^[5]. He presents that rule of thumb as the status quo he is criticizing, not as a finding. The same deck calls appraisers' practice "an illiquidity bludgeon" applied "with little or no differentiation across businesses," and says the restricted-stock and IPO studies used to justify it "have small samples" with "significant sampling bias."

Drivers: size, liquidity, concentration, owner dependence

No single dataset decomposes the discount. The sources above name the mechanisms, and each one points to why a smaller company tends to clear at a lower multiple.

Liquidity varies, and it should. Damodaran argues illiquidity "should vary across" companies, time and buyers: "Healthier and larger companies, with more liquid assets, should have smaller discounts than money-losing smaller businesses with more illiquid assets"; liquidity "is worth more when the economy is doing badly and credit is tough to come by"; and it is worth more to buyers "who have shorter time horizons and greater cash needs" ^[5]. A flat 20-30% haircut ignores all three.

Undiversified buyers. In a private-to-private sale, Damodaran notes, "Neither the buyer nor the seller is diversified," so models that price only non-diversifiable risk "will seriously underestimate the discount rates" ^[5]. An individual buyer putting most of their net worth into one business demands a higher return, which means a lower price for the same earnings.

Owner dependence. The same slides list "Key person value": revenues and operating profit "reflect not just the potential of the business but the presence of the current owner," and "the life of a business that is built around a person or even a family can be constrained" ^[5]. Smaller companies are more often run by an owner-operator who is also the main salesperson, estimator and customer relationship. BizBuySell's Q3 report says businesses with "limited owner dependence continue to attract significant attention" ^[2].

Concentration. A small company tends to depend on fewer customers, suppliers and people. None of the datasets above measures concentration, so we do not quote a figure, but the mechanism is the same as owner dependence: losing one relationship changes the earnings the multiple is applied to.

Financing access and buyer pool. BizBuySell reports that smaller transactions "remain heavily dependent on SBA funding," which brokers named "the most common financing obstacle of the quarter, cited by 38%," and that larger businesses "draw buyers less dependent on SBA" ^[2]. The IBBA release makes the competition point directly: deals under $500,000 "frequently receive just one or two offers," while "87% of deals over $5 million attracted at least three offers, with 33% attracting an impressive 10+ bids" ^[3]. IBBA's 2026 chairman summarized it this way: "At the smallest end of the market, however, buyers have more leverage and are more sensitive to financing, margins, and operating risk" ^[1].

Size premia are contested. Damodaran's slides also say the research behind the "small cap" premium in build-up discount rates is "Dated (The small cap premium has not existed since 1981)" ^[5]. That is one author's position, not a settled finding, but it is a reminder that "small companies are riskier, so they are worth less" is an argument to test, not a law.

What appraisers and investors report

The Pepperdine Private Capital Markets Project surveys lenders, investors, appraisers and brokers each year. Figures below are read from its 2026 report (survey deployed January 2026), which is free to download ^[6].

• Appraisers (211 respondents). The median market (equity) risk premium is 6.00%. The overall median discount for lack of marketability (DLOM) is 15%. Both fall with size: median DLOM for a controlling interest declines "from 12.0% for companies with $100K in revenues to 7.0% for companies with $250M in revenues" (minority interests: 30.0% to 20.0%), and the median company-specific risk premium is 5.00% at $1M of revenue against 2.00% at $250M.
• Private equity (38 respondents). Median required returns are 20% for a $1M-EBITDA company, 25% at $5M and $10M, and 19% at $50M ($25M had too few responses) (Table 1). On buyouts actually completed, median expected IRR is 20% in the $1M-$4.99M EBITDA band and 21-23% from $5M to $49.99M (Table 40).

The appraiser figures point the same way as the broker data: smaller companies carry a larger marketability discount and company-specific premium, so a lower price for the same earnings. The PE figures do not: required returns peak at $5M-$10M of EBITDA, and with 38 respondents and some size cells resting on two to four deals, that is too thin to read as a size effect. All of these are survey medians of stated assumptions, not transaction prices, and a DLOM is an appraisal input, not the gap between two transaction multiples.

How to read a published range: a five-question checklist

Before a multiple goes into a model, a memo or a negotiation, answer five questions about it.

Illustrative example (fictional). Two newsletters land on a search fund principal's desk. One says "companies like this sell for a median of 6x." The other says "this industry trades at a median of 3x." The principal is looking at a fictional HVAC services company, Northfield Mechanical, with these numbers:

At "3x" on SDE, the implied price is $3.0 million. At "6x" on EBITDA, it is $5.1 million. The two claims look contradictory, but they may describe different buyers, different sizes and different periods, and neither newsletter said which earnings it used or how many deals sat behind its median. Neither number tells the principal what Northfield is worth. The comparison only shows that the two figures cannot be placed on the same axis until all five questions are answered.

This is a description of how published datasets are constructed, not valuation advice.

Where the owner population sits

Most acquisition targets are small because most companies are small. The SBA Office of Advocacy's 2026 FAQ reports that "There are 36,207,130 small businesses in the United States" and that "99.9 percent of businesses are small" ^[7]. That figure is not a count of acquirable companies. It is the base from which every target list, including one aimed at the lower middle market, is drawn.

The value distribution is highly skewed. The Federal Reserve's Survey of Consumer Finances, published in October 2026, found that 15.4% of families held private business equity in 2025, with a conditional median value of $75,400 and a conditional mean of $1,566,600 ^[8]. The Fed's own explanation: "The large differences between the median and mean values reflect the small fraction of privately held businesses with very high valuations."

For anyone screening targets, that skew has two consequences. First, the typical owner's business is far below the size at which the IBBA ladder reaches its higher rungs. Second, averages computed over mixed populations, like the market-wide figures in any broker report, are pulled by the few large deals in the sample.

What this means for screening

The size gradient in multiples is real in direction and unreliable in magnitude. For a deal team building a target list, that suggests using size as a screening criterion without attaching a price to it.

• Write size bands in the earnings measure you will underwrite. If the fund underwrites EBITDA after a replacement manager, state the floor that way in your screening criteria, not as SDE or revenue.
• Treat owner dependence as a question to answer, not a discount to assume. Who holds the customer relationships, who prices the work, who signs the bonding or licenses: these are checkable facts about a target, and they bear directly on the drivers above.
• Ask about concentration early. The number of customers above a set share of revenue is cheaper to learn before a call than after an LOI.
• Keep every published multiple attached to its five answers. A range without its earnings basis, sample, band, period and structure belongs in a footnote, not a model.
• Record what you could not verify. A gap that stays visible is more useful than an estimate filled in to make a row look complete.

Docket runs this kind of evidence check on target lists before outreach; see our evidence standards.