Private equity's middle market shifted its weight toward smaller companies in the second quarter, and the numbers are not subtle. PitchBook's latest US PE Middle Market Report counts 413 sponsor deals valued between $25 million and $100 million in the quarter, up 56.4 percent from the first quarter, with combined value of $16 billion, up 70.6 percent. It was the only size band that grew on both measures. For owners of companies that fit inside that range, the buyer universe just got larger, and the way those buyers underwrite a founder-owned business has changed with it.
The migration shows up at both ends of the size spectrum. At the top, only eight upper-middle-market deals (enterprise values between $500 million and $1 billion) were signed in the quarter, down about 64 percent, with total value falling to $5.7 billion. At the bottom, the $25 million to $100 million band posted its 56 percent jump in count. The median middle-market deal fell to $151.9 million in the first half of 2026 from $179 million in 2025.
Total middle-market deal value still slipped 8.7 percent to $92.4 billion, but deal count rose 25.7 percent to 1,090. Sponsors are deploying capital; they are simply dividing it into smaller pieces.
The fundraising side confirms the direction. Siguler Guff closed its sixth small-buyout fund on September 17 at more than $3 billion, its largest ever for the strategy and above its $2.2 billion target. The firm describes its average portfolio company as family-owned, more than 30 years old, and new to institutional capital, with EBITDA typically between $5 million and $50 million. That describes a large share of the privately held companies in the United States.
The more important change is in the playbook, not the volume. For most of the past decade, a sponsor's standard move was to buy a scaled, integrated platform and bolt smaller peers onto it. Founder-owned companies in the $25 million to $100 million range were the add-ons: bought at lower multiples, folded into someone else's platform, and often paid partly in that platform's equity.
Advisers quoted in the PitchBook report describe sponsors now buying those same small, founder- or family-owned businesses as the platform itself, with an explicit plan to grow through acquisitions from there. Fragmented sectors with heavy back-office overlap are where this is concentrated: industrial services, insurance brokerage, and residential services.
The reason is supply. Fewer sponsors are selling their portfolio companies (middle-market exit value fell 19.5 percent in the quarter to $24.7 billion, the lowest since the second quarter of 2020), so the pool of scaled platforms a buyer can purchase is thin, and the few that surface command full prices. A buyer that cannot find a $400 million platform at a sensible multiple can instead build one from a $60 million company and a pipeline of $10 million add-ons.
For an owner, the distinction cuts both ways. A platform buyer is paying for the management team, systems, and market position that will carry future acquisitions, so the seller's leverage on price, rollover terms, and post-close role is stronger than an add-on seller's. But that buyer is also underwriting a plan that depends on the business absorbing other companies, so diligence goes deeper on what makes integration possible: financial reporting, customer concentration, the second layer of management, and whether the founder can step back without the business stepping back too.
Pricing by size band explains the shift and what an owner should expect. PitchBook's first-quarter data put the median entry multiple for companies with enterprise values between $500 million and $1 billion at 13.2 times EBITDA, up from 12.1 times in 2025. For companies in the $25 million to $100 million range, the median was 8.5 times, down from 8.8 times a year earlier. Advisers put the smallest businesses lower still, in the range of four to eight times EBITDA.
That gap is the entire thesis. A sponsor that buys at 8.5 times, adds several companies at six times, and sells the combined business at 11 or 12 times has made money on the arbitrage alone, before any operating improvement. The lower entry multiple on a founder-owned company is not a discount for quality. It is the price of the integration work the buyer is signing up to do.
Two things follow for a seller. First, the multiple you are offered will be benchmarked against the size band you are in, not the size band you might be in after the buyer's plan works. Arguing for 11 times because a scaled competitor sold for 11 times rarely wins on its own. Second, the way to capture part of the arbitrage is structural: rollover equity in the platform, an earnout tied to the platform's growth, or a role in the acquisition program itself. Each puts the seller on the same side of the multiple expansion as the buyer.
Leverage also behaves differently at this size. Siguler Guff reports that its deals typically carry debt of two to three times EBITDA, against five to six times in larger buyouts. With the Federal Reserve having raised rates on September 16 to a range of 3.75 to 4.00 percent, that lighter debt load is part of why smaller deals kept closing while larger ones stalled. It also means the buyer's return depends more on growth and integration than on financial engineering.

Headline multiples are the starting point; the structure is where the money moves. PitchBook notes that deals are taking longer to close because buyers and sellers disagree on price, and that dealmakers are leaning on earnouts (payments contingent on future performance) to bridge the gap. The lower middle market already uses them heavily: SRS Acquiom's 2026 study of private-target deals found earnouts in 35 percent of transactions under $25 million and 29 percent of all deals under $50 million.
Escrows follow the same pattern. The same study found 61 percent of lower-middle-market deals carried a general indemnification escrow (a portion of the purchase price held back to cover breaches of the seller's representations), and 43 percent carried more than one. The difference between the headline price and cash at closing can easily be 10 to 20 percent of the deal, with part of it contingent on results the seller no longer controls.
A platform buyer will also press on rollover. A seller who rolls 20 to 30 percent of proceeds into the new platform is taking a second bet on the buyer's acquisition program: the most valuable piece of the deal if the program works, and the least liquid if it does not. The terms that govern that stake (the class of equity, whether it sits behind the sponsor's preferred return, what happens on a sale or recapitalization) deserve as much attention as the cash price.
Three things will shape whether this shift holds. The exit market is the first: middle-market exits at their lowest level since 2020 are what pushed buyers down-market, and a reopening of the sale market for scaled platforms would pull some of that capital back up. The second is the November midterm elections. Advisers in the PitchBook report flag healthcare and youth sports as sectors that could face more scrutiny if control of the House changes, and expect deal flow to pick up once the outcome is known. The third is fundraising concentration. Middle-market funds raised $69.1 billion in the first half, on pace for roughly $140 billion for the year, but the number of funds is set to decline for a third straight year: fewer, larger managers with dedicated small-buyout mandates, and a more organized buyer set for founder-owned companies.
The buyer most likely to call a company in the $25 million to $100 million range this year is a sponsor that wants to build a platform, not just fill one. That buyer will pay a multiple set by size band, expect to earn the gap to larger-company multiples through integration, and offer the seller a share of that gap through rollover and earnouts rather than cash. Owners who prepare for that conversation, by knowing their band, cleaning up the things a platform buyer tests first, and modeling the structure rather than the headline, will get more of the value the buyer is planning to create. Those who wait for the call will get the buyer's version of the plan.