Private Equity Multiple Compression Watch for 2026

Private equity multiple compression in 2026 is selective rather than uniform. CohnReznick puts median implied EV/EBITDA multiples in US private equity at 13.4x in the first half of 2026, down from 14.6x in 2025. That is a real move, yet it sits alongside a Bain and StepStone survey, cited by Bain’s Hugh MacArthur, showing that 79% of GPs expected purchase price multiples to stay flat this year. Both can be true. The market is separating assets that can defend a premium from assets that cannot, while the return model that carried the last cycle, cheap debt plus multiple expansion, has stopped working as a base case.

Market signals moving in 2026

The clearest read on the year comes from signals that pull in different directions. Headline multiples are lower, GP expectations are largely flat, entry prices remain historically high, leverage contributes less to returns and exits remain constrained. In practice, that combination leaves little room for underwriting error.

The divergence between count and size is important. CohnReznick puts median target EBITDA at $30.3m in Q1 2026 and $64.5m in Q2, against a 2025 full-year average near $30m. Capital is concentrating in larger platforms, which is itself a valuation statement because scale, resilience and exitability are attracting a larger share of available capital.

Entry pressure, exit pressure and marks

Multiple compression means investors pay less for each unit of earnings. Investopedia frames it as a public-market phenomenon driven by rising rates, slowing growth or weaker sentiment. In buyouts the relevant metric is usually enterprise value divided by EBITDA, and the compression question splits four ways.

• Purchase price multiple compression: sellers accept lower entry multiples, which helps new deployment.
• Exit multiple compression: the multiple realised on sale falls below the multiple paid, which destroys equity value in existing holdings.
• Mark-downs: carrying values are reduced in response to comps or transaction evidence, with no cash consequence until exit.
• Public comps compression: the reference set moves, forcing valuation committees to justify any widening gap.

A sponsor can grow EBITDA by a third over a hold and still return less than cost if the exit multiple gives back two turns. That arithmetic explains why exit multiple expansion was such a powerful and under-examined return driver in the previous cycle. Removing it changes the threshold for a credible investment case.

Flat multiples still create pressure

McKinsey’s median purchase multiple of 11.8x in 2025, against a 2010 to 2022 average of 9.1x, makes the flat-multiple case uncomfortable. Holding an elevated entry multiple constant through exit is not a benign assumption when the starting point sits nearly three turns above the long-run average.

Leverage no longer fills the gap. McKinsey puts debt at 37% of entry multiples, seven percentage points below the 2010 to 2022 average of 44%. Sponsors are writing larger equity cheques into higher-priced assets, so the equity return depends more heavily on operating performance and less on the capital structure.

MacArthur summarises Bain’s version of the maths bluntly: deals that once needed roughly 5% annual EBITDA growth to produce a 2.5x now need closer to 10% to 12%. His shorthand is “12 is the new 5.”

The return bridge under flat multiples

Take a business bought at 12.0x EBITDA of $50m, so $600m enterprise value, funded with 40% debt of $240m and $360m equity. Assume a five-year hold, exit at the same 12.0x, and $100m of cumulative debt paydown from free cash flow.

• At 5% annual EBITDA growth, EBITDA reaches roughly $64m. Exit EV of $766m less $140m net debt gives $626m equity, about 1.7x, or roughly 12% IRR.
• At 12% annual growth, EBITDA reaches roughly $88m. Exit EV of $1,057m less $140m net debt gives $917m equity, about 2.5x, or roughly 21% IRR.
• Drop the exit to 11.0x with 12% growth and equity falls to roughly $829m, about 2.3x.

The gap between the first and second line is entirely operating performance. One turn of compression on the third line costs roughly 0.2x of MOIC, which is survivable at 12% growth and terminal at 5%. Build these as separate cases rather than sensitivities buried in a terminal value cell. Standard LBO modelling practice should now run flat, minus 0.5x and minus 1.0x exit cases as defaults.

Assets still clearing at a premium

CohnReznick’s dispersion finding is the most commercially useful data point of the half. Premium pricing survives where demand is contracted, regulated or non-discretionary. Pressure lands on anything cyclical, discretionary or exposed to input volatility.

BDO’s read is consistent: valuations remain high for high-quality assets because capital competes for a limited pool of resilient, strategically essential companies. Scarcity, rather than broad sentiment, is holding those multiples up.

PwC flags AI repricing as one of two watch items for the second half of 2026, alongside the LP liquidity imperative. The effect runs both ways. AI compresses terminal value assumptions for software businesses whose moat was workflow lock-in, while it raises the plausible margin ceiling for services businesses that can automate delivery. Treat it as a terminal value input, rather than a growth narrative.

The exit bottleneck is where marks meet cash

A multiple is a hypothesis until something clears. CohnReznick counted 872 US PE-backed exits in the first half of 2026 against 1,210 in the same period of 2025, while PwC describes exit activity as suppressed, with dispersion widening between firms that can show realised returns and those that cannot.

Simmons & Simmons, writing in January 2026, notes that the exit backlog is larger in value, count and share of portfolio companies than at any point in two decades, across more than 30,000 PE-backed companies globally. McKinsey’s 6.5-year average hold and 16,000-company four-year-plus inventory are the same constraint measured differently.

Megadeals coexist with this pressure. McKinsey records the announced $55bn take-private of Electronic Arts in 2025 as the largest PE deal in history, with buyout and growth deals above $500m up 44% to more than $1tn. Headline deal value and difficult underlying return maths can sit in the same market.

Continuation vehicles solve liquidity, not price discovery

PwC identifies continuation vehicles and secondaries as the primary liquidity release valve in 2026. Simmons & Simmons, citing Raymond James, reports that roughly one-fifth of PE sales in 2025 involved groups raising money from new investors to acquire businesses from older funds, up from 12% to 13%, worth around $105bn, with 2026 expected to set another record.

The valuation conflict is structural. The selling fund wants a price that validates its carrying value. Incoming investors want entry-level upside. Continuing LPs want a defensible reference price. Exiting LPs want cash. A continuation vehicle priced to the mark is a weak test of that mark, which explains why LPAC scrutiny of these processes has tightened.

PwC reports LPs demanding realised returns over paper marks, with DPI becoming a defining fundraising metric in 2026. Aggregate fundraising dollars rose 9% in the first half against fewer fund closings, with capital concentrating among top performers. McKinsey’s survey of 300 LPs found around 70% planned to maintain or increase PE deployment in the second half, against 68% in the first. Allocations are not retreating. They are consolidating around managers that can convert marks into distributions.

Re-underwriting priorities before year-end

Segment the portfolio by liquidity path rather than by sector or vintage.

1. Exit-ready premium assets: recurring revenue or healthcare exposure with a live buyer set. Test the market while pricing supports it.
2. Hold and improve: assets where a defined operating plan can deliver the 10% to 12% growth the flat-multiple case requires.
3. Valuation risk: cyclical, discretionary or AI-exposed holdings where the current mark assumes an exit multiple the 2026 market will not pay.
4. Structured liquidity candidates: businesses suited to a minority recap, GP-led secondary or NAV facility where a full exit is not available at an acceptable price.

For live processes, monitor sponsor-to-sponsor bid-ask spreads, continuation vehicle pricing relative to carrying value, exit volume rather than exit value, and private credit terms including spreads, original issue discount, covenant packages and payment-in-kind usage. BDO expects deal volume and value to climb in 2026 as borrowing costs decline and tariff uncertainty recedes, which would ease the financing constraint without restoring multiple expansion.

Every operating initiative in the base case should be named, timed and quantified before investment committee approval. Value creation plans that were sensible additions to a leverage-driven return are now the return itself.

Conclusion

The decision rule for 2026 is straightforward. If a deal only works with exit multiple expansion, it does not work. Underwrite flat as the base case, model minus one turn as a live scenario, and require the EBITDA growth that Bain’s arithmetic implies before approving the equity cheque.

Getting this wrong does not produce an immediate loss. It produces a portfolio company held for seven years, marked at a multiple no buyer will pay, funded through a NAV facility while the fund’s DPI stays near zero and the next raise stalls. The cost shows up in the following fundraise, not the current model.

P.S. If entry and exit multiples are shaping your underwriting, check out our Premium Resources for LBO models, PE & VC databases and more tools to help you advance your career.

Sources

• Private Equity Mid-Year 2026 Trend Report
• Global Private Equity Report 2026
• Private equity: US Deals 2026 midyear outlook
• Private Equity Outlook 2026: Flat Multiples, Higher Earnings Growth
• Private Equity – The Year Ahead
• Private Equity Outlook 2026
• Multiple Compression: Definition and Insights for Investors