Executive Summary
The composition of private equity returns has inverted between 2022 and 2026. Multiple expansion and leverage together produced roughly 59% of PE returns from 2010 through 2022. By 2026, that combined contribution has fallen to less than 35%, with multiple expansion specifically at 32% and trending lower as 80% of GPs now model flat or compressed exit multiples. Revenue growth has stepped in as the dominant return source at 54%, with margin improvement at 14%. The challenge facing the industry is not understanding this shift in the abstract. It is reconciling underwriting models that still assume 10 to 12% portfolio company EBITDA growth with realized portfolio company growth of 3.8% in 2024. The gap is wide enough that a meaningful share of the 2020 and 2021 vintages were mispriced at the asset level on day one.
The Return Composition That Defined the Era
For a generation of PE professionals trained between 2010 and 2022, the return decomposition was almost a constant. A typical buyout return broke down into roughly the following contributors: multiple expansion at 27%, leverage and capital structure at 32%, revenue growth at 24%, and margin improvement at 17%. Variations existed across sectors and strategies, but the dominant return drivers were the two contributors that depended least on the actual business performance: multiple expansion and leverage.
The structural reasons for that pattern are well understood in retrospect.
Multiple expansion was a function of falling interest rates, growing PE dry powder competing for assets, and a strong IPO and strategic exit market. From 2010 to 2021, average exit multiples expanded by roughly four to five turns across most buyout sectors. Software, healthcare, and certain consumer categories expanded more. The compounding effect across a typical five-year hold turned a 0.5 to 1.0x multiple expansion into 15 to 25% of total return.
Leverage and capital structure benefits flowed from cheap and abundant debt. Acquisition financing at 4 to 5% all-in costs, refinancing windows that allowed sponsors to recapitalize at lower rates, and dividend recapitalizations that returned capital before exit all contributed. The leverage tail also amplified equity returns mechanically: a 6x debt-to-EBITDA capital structure converts modest operational improvement into outsized equity multiples.
Revenue growth and margin improvement were genuine contributors but typically the smaller portion of the return story. The PE operating thesis added value, but the financial backdrop produced most of the returns.
This composition created a generation of investment professionals whose training emphasized capital structure optimization, exit timing, and deal selection in markets with reliable tailwinds. Operational alpha was a topic in case studies and at conferences, but the dollars that flowed to LPs came primarily from financial physics.
What Disappeared Between 2022 and 2024
Two of the four return contributors disappeared in an 18-month window.
The rate environment shifted starting in early 2022. Federal Reserve action moved the effective federal funds rate from 0.25% to roughly 5.25% in less than two years. Spreads on PE acquisition debt widened on top of the base rate move. By 2024, the all-in cost of leverage for a typical sponsor LBO sat at 8 to 10% on senior debt and well into the teens on unitranche or junior capital. Refinancing windows that used to be opportunistic became hostile. Dividend recapitalizations slowed sharply. The capital structure contribution to returns, historically 30%+, dropped to single digits across most vintages.
The exit multiple environment compressed shortly after. Public market multiples in software and growth sectors compressed by 40 to 60% from 2021 peaks. Industrial multiples held up better but still compressed. Strategic buyers, facing their own cost of capital pressure, became more disciplined on premium pricing. Sponsor-to-sponsor exits, traditionally the safety valve, slowed when buying sponsors could not justify the multiples that selling sponsors required. By the time the 2022 and 2023 GP surveys were completed, roughly 80% of respondents were modeling flat or compressed multiples at exit for new investments. The multiple expansion contribution to expected returns dropped from the historical 27% to roughly 32% in modeled assumptions, but with the directional bias trending lower, not higher.
The combined impact eliminated more than half of the historical return contribution. What had been a 59% structural tailwind in the 2010 to 2022 era became a 5 to 15% contributor in modeled returns for new investments, with downside skew if multiples compress further.
The New Composition
By 2026, the typical buyout return composition has rearranged into something close to the following: revenue growth at 54%, multiple expansion at 32%, margin improvement at 14%. Leverage as a return driver has effectively collapsed into a neutral or modest negative contributor depending on the deal.
Three observations matter about this new composition.
The first is that 54% revenue growth as a return contributor implies meaningful portfolio company top-line performance. Modeling a five-year hold with revenue growth driving more than half of total return requires either substantial organic acceleration in portfolio companies or programmatic M&A that compounds revenue inorganically. Both paths require operational infrastructure that mid-market PE has historically not built at scale.
The second is that the 32% multiple expansion contribution to underwritten returns is in tension with the 80% of GPs modeling flat or compressed exit multiples. Either the modeled returns are systematically optimistic on the exit multiple, or GPs are assuming that the specific companies they own will defy the sector trend through improved positioning, scale, or operational excellence. The aggregation suggests that some portion of underwritten returns will not materialize at exit multiples.
The third is that margin improvement at 14% is meaningful but bounded. Most portfolio companies cannot expand margin by 500 to 1,000 basis points in a five-year hold. The companies that can are typically in early-life operating maturity where the playbook is well-known: pricing optimization, procurement consolidation, headcount efficiency, and SG&A leverage. Margin improvement carries return contribution but it is not the lever that compensates for lost multiple expansion at scale.
The composition matters because it dictates which operational capabilities GPs need to build. Revenue growth at 54% implies investment in sales capacity, pricing, channel expansion, and M&A integration. Margin at 14% implies the operational efficiency tooling that PE has historically focused on. The combined demand is for a fund-level operating infrastructure that mid-market generalist firms have been slower to develop than mega-funds.
For a deeper look at how the largest PE firms have responded to this shift, see operational value creation in PE.
The Underwriting Gap
The structural inversion is widely recognized at the industry level. At the deal level, it has not fully propagated.
Bain's 2026 outlook and corroborating data from McKinsey and Apollo's research desk all point to the same gap. Portfolio companies in PE-backed sectors achieved roughly 3.8% EBITDA growth in 2024. Deals being underwritten in 2025 and 2026 are typically modeled at 10 to 12% annual EBITDA growth across the five-year hold. The gap of roughly seven percentage points compounded over five years represents the difference between a deal that returns 2.5x at the underwriting assumption and a deal that returns barely 1.5x at the realized growth rate.
The underwriting gap exists for understandable reasons. GPs need to underwrite deals at returns that clear LP hurdles. The arithmetic of clearing a 2.5x return at a 15 to 17x entry multiple with no leverage tailwind and flat exit multiples requires either operational improvement that lifts EBITDA significantly or growth assumptions that exceed historical realized growth. Most deals have chosen the growth assumption because it is the more straightforward path through the underwriting committee.
The consequence is that a meaningful share of deals priced in 2024 and 2025 were mispriced at acquisition. The portfolio companies will not deliver the modeled growth. The exit multiples will not expand to bridge the gap. The deals will return capital below the underwriting assumption, in some cases substantially below.
The cohort effect across the industry will surface gradually. Most of these deals are in year one or year two of holds. The gap between underwriting and execution becomes visible in year three when annual operating reviews show realized performance below model. By year four, GPs are calibrating exit timing assumptions downward. By year five, the realized exit multiples are setting market clearing prices that subsequent GPs use to underwrite their own deals.
The cycle has historically taken three to five years to clear. The current cycle is roughly mid-way through that process. Vintages priced in 2026 will benefit from the realism that the 2020 and 2021 vintages did not have. Vintages from 2022 through 2024 will work through the gap on the way to realization.
The Two GP Paths
Faced with the underwriting gap, GPs have two reasonable paths. Both have meaningful problems.
The first path is extending hold periods. If the realized growth comes in slower than underwritten, hold the portfolio companies longer until the realized growth produces the modeled EBITDA. The arithmetic works. A company underwritten to reach $80M EBITDA in year five but only reaching $60M in year five can be held to year seven or year eight, with continued operational work, to bridge the gap.
The problem with extending holds is twofold. The first is that IRR compresses sharply with extended duration. A 2.5x return at year five produces roughly 20% IRR. The same 2.5x at year eight produces 12%. LPs underwriting at 18 to 20% net IRR get returns closer to 10 to 12% net. The IRR-to-DPI relationship that defines fund performance worsens. The second is that extended holds compound the existing distribution drought. LPs have already been waiting longer than expected for distributions. Asking them to wait another two to three years compounds the credibility problem.
The second path is repricing. Entry multiples reset lower to reflect realistic growth assumptions. New deals get priced at 10 to 12x rather than 15 to 18x. The arithmetic of the deal works because the entry multiple is lower, the implicit exit assumption is more conservative, and the return clears LP hurdles without requiring 12% portfolio company growth.
The problem with repricing is that it is happening too slowly. Sellers, including PE sponsors selling existing portfolio companies, are reluctant to accept the new multiples. Limited supply of high-quality assets at the reset multiple means the few deals that do get done at lower entry prices are highly competitive. Most processes still close at multiples that imply growth assumptions that historical data suggests will not materialize.
Most GPs in practice are running both paths simultaneously: extending holds on existing portfolio companies while attempting to underwrite new deals at more disciplined multiples. The combination is producing a slower deal pace overall, with fewer transactions closing than the dry powder would suggest.
What This Means for LP Allocation
LPs are reading the return composition shift in their allocation decisions. The signals show up in several places.
First-fund and second-fund allocations are increasingly going to GPs with documented operational capabilities. The LP underwriting question has moved from "what is your track record" to "what is your operational infrastructure" because the track record from recent vintages will not differentiate enough on its own. GPs with credible operating partner teams, sector specialization, and value creation playbooks are getting through allocation committees faster than generalist mid-market firms with thin operating layers.
Re-up decisions for established managers are scrutinizing portfolio company growth and margin trajectories more carefully. LPs want to see specific evidence of operational alpha in the current portfolio, not just claims about the historical playbook. Portfolio company KPI reporting, the kind that surfaces revenue growth, customer cohort behavior, and margin trends in real time, has become a differentiator in re-up conversations.
Strategy rotation toward secondaries, co-investment, and direct sponsor vehicles continues. LPs that are uncertain about the return profile of new primary commitments are deploying through structures that offer either clearer asset-level visibility or shorter expected hold periods. Secondaries fundraising hit $39 billion in Q1 2026 alone, the second-highest first quarter on record.
Sector allocation is shifting toward areas with structurally stronger growth dynamics. Industrial services, business services with embedded software, healthcare services, and certain defense and infrastructure subsectors are attracting allocation faster than consumer or generalist software. The underlying logic is that revenue growth, the dominant new return contributor, requires sectors where 5 to 10% organic growth is realistic.
For more on how this connects to fundraising strategy, see what LPs actually want in 2026.
What Has to Change in Underwriting
For the structural inversion in returns to be reflected in deal pricing, three changes need to propagate through underwriting practices.
The first is growth assumption discipline. Underwriting that assumes 10 to 12% portfolio company EBITDA growth needs to be either downgraded to realistic levels or supported with very specific operational levers that explain the gap. A deal modeled at 12% growth in a sector with 4% organic growth needs to identify which 8 percentage points come from operational improvement, M&A, pricing, or share gain, with specific evidence from comparable deals or portfolio companies.
The second is multiple assumption discipline. Underwriting that assumes flat or modest expansion in exit multiples requires acknowledgment that 80% of GPs assume the same. If everyone is modeling flat multiples, the realized multiples will be set by the market clearing dynamics, not by the modeled assumption. Some sectors will compress further. Some will hold flat. A small number may expand. Underwriting needs to handicap which category each specific deal falls into.
The third is downside scenario weighting. The historical practice of underwriting to a base case with modest downside sensitivity needs to be replaced with more rigorous downside scenarios that reflect the realized macro environment. What does the deal return if revenue growth comes in at 4% rather than 10%. What does it return if exit multiples compress by another two turns. What does it return if rates remain at current levels through 2030. The downside scenarios are no longer hypothetical edge cases. They are the realistic environment that recent vintages are working through.
GPs that build this discipline into their underwriting will close fewer deals at lower multiples and return capital more reliably. GPs that continue to underwrite at the historical assumptions will close more deals at the existing multiples and return capital below the underwriting in a meaningful share of cases.