Executive Summary
Nine consecutive PE vintages have failed to deliver median DPI above 1x. The 2016 cohort was the last to reach that threshold. The math has been deteriorating since 2017, accelerated by an exit market that contracted in 2022 and never fully reopened. Median hold periods have stretched to 6.4 years, only a handful of buyout IPOs have priced in recent cohorts, and roughly 61% of exit value across the 2017 through 2022 vintages has come from sponsor-to-sponsor secondaries and GP-led continuation vehicles. The gap between reported TVPI and realized DPI has reached the widest point in modern PE history. LPs have stopped accepting unrealized marks as evidence of performance and are filtering re-up decisions through a single question: when do I get my money back.
The Vintage That Drew the Line
The 2016 vintage was the last to behave the way mature PE math is supposed to behave. Funds raised in that vintage entered the market when entry multiples sat at roughly 9 to 10x EBITDA, leverage was abundant at 4 to 5% all-in cost, and the exit window from 2019 through 2021 produced both strong strategic activity and a meaningful IPO calendar. The result was median DPI above 1.0x by year seven, with top-quartile DPI well above 1.5x.
The 2017 vintage looked similar on entry but ran into the back half of the cycle as the exit market began to compress. Median DPI for 2017 sits at roughly 0.85x at the same point in lifecycle. 2018 sits at 0.75x. 2019 at 0.55x. 2020 at 0.35x. The 2021 vintage, which represents the largest single dollar deployment in the history of the asset class, is tracking at median DPI below 0.20x as of early 2026. The 2022 and 2023 vintages are still too young to evaluate with conviction, but the underlying portfolio companies were acquired at multiples that have since compressed, with debt that has since become more expensive to service.
The pattern is consistent. Each successive vintage from 2017 forward has delivered lower realized cash returns at the same point in fund lifecycle than the vintage before it. The compounding effect is significant. LPs holding diversified PE portfolios across these vintages have been receiving distributions stuck at 14 to 15% of NAV since late 2022. That distribution rate sits roughly half of the long-run norm of 25 to 30%.
For context, the previous distribution drought, 2008 through 2010, lasted six to eight quarters. The current drought is now in its fifteenth consecutive quarter.
What Happened Between TVPI and DPI
Total value to paid-in capital tells a different story than DPI right now. The average North American buyout fund from the 2018 through 2021 vintages reports median TVPI between 1.6x and 1.9x, depending on the data source and the vintage. On paper, those numbers look acceptable. Below 2.0x but not catastrophic.
The disaggregation tells the real story. A typical 1.8x TVPI on the 2019 vintage breaks down to roughly 0.55x DPI plus 1.25x unrealized RVPI. The cash returned to LPs is just over half of paid-in capital. The remaining 125% of paid-in capital exists in the form of valuation marks on portfolio companies that have not exited, have not been refinanced at scale, and in many cases are sitting on capital structures that were sized for a different rate environment.
The gap between TVPI and DPI is structurally meaningful for three reasons.
The first is the auditability of the mark. Public market comparables have compressed sharply since 2021. Software multiples have gone from 12 to 15x revenue at the 2021 peak to 5 to 8x revenue in 2026. Industrial multiples have compressed less but compressed nonetheless. PE valuation methodology uses public comparables as one input, but the marks have not fully reflected the compression. Auditors have applied increasing pressure, and LPs have noticed. The unrealized portion of TVPI now carries a credibility discount that did not exist in 2021.
The second is the financing of unrealized value. A meaningful portion of recent vintage portfolios is held in capital structures with debt costs that exceed the cash flow yield on the underlying business. These positions cannot be marked up. They cannot easily be sold. The only paths to realization are restructuring, lender takeover, or a multi-year hold while the business grows into its capital structure. None of those paths preserve the IRR underwritten at acquisition.
The third is the dependence on financial engineering exits. GP-led secondaries, continuation vehicles, NAV loans, and dividend recaps have all increased substantially since 2022. Each of these mechanisms delivers some form of capital movement to LPs without representing a true exit. A continuation vehicle that rolls a portfolio company into a new fund structure crystallizes the mark but does not return the underlying asset to liquidity. A NAV loan delivers cash to LPs but represents leverage on top of an already-levered position. These tools have legitimate uses, but their widespread adoption since 2022 has muddied the line between value creation and financial restructuring.
Why Hold Periods Are Stretching
The natural consequence of a compressed exit market is longer hold periods. The median PE hold has moved from 4.5 years in 2017 to 6.4 years in 2025. The top decile of holds, the assets that GPs are most reluctant to sell at current valuations, sits above eight years.
Three exit channels have all contracted at the same time.
Strategic M&A has slowed. Corporate buyers facing higher capital costs, shareholder pressure on integration risk, and antitrust scrutiny have been less aggressive on premium-priced sponsor exits. Strategic-to-sponsor flow as a portion of buyout exits has dropped from roughly 40% in 2019 to 25 to 30% in 2025.
IPOs have remained intermittent. The PE-backed IPO calendar has been thin since 2021, with windows opening briefly and then closing. Recent vintages of large PE portfolios have produced only a handful of IPO exits, most concentrated in software and healthcare assets that could clear public market discipline. The bulk of large portfolio companies in industrial services, business services, and consumer remain private.
Sponsor-to-sponsor exits, historically the safety valve, have compressed at the seller end. The same multiple compression that hurts the seller hurts the buyer. Buying sponsors have been less willing to pay the multiples that selling sponsors require to clear underwriting hurdles. The gap between bid and ask has been wide enough in many sectors that processes are launched, marketed, and then withdrawn.
The fourth channel, GP-led secondaries and continuation vehicles, has grown to fill the gap. Secondaries fundraising in Q1 2026 hit $39 billion, the second-strongest first quarter on record. GP-led secondaries, in particular continuation vehicles, now represent the largest single category within secondaries flow. These transactions deliver some liquidity to LPs who exit and rolling exposure to LPs who continue, but they do not represent organic value crystallization. The cohort effect is that exit value attribution has shifted: roughly 61% of exit value across recent vintages now flows through some form of sponsor-managed liquidity mechanism rather than third-party realization.
The longer holds have IRR consequences. A 5x return delivered in five years produces a roughly 38% IRR. The same 5x return delivered in eight years produces 22%. LPs that underwrote 2019 commitments at 20%+ net IRR are now modeling realized returns closer to 12 to 15% net.
The Zombie Portfolio Problem
Beneath the distribution drought sits a structural issue that GPs are reluctant to name directly. A meaningful portion of recent vintage portfolios consists of companies that are technically solvent, technically performing, but not generating realistic exit paths at marks consistent with the underwriting.
The pattern is most visible in PE-backed software acquired between 2020 and 2022. Many of those companies were purchased at 12 to 18x ARR multiples, with growth assumptions of 25 to 35% annually. Actual revenue growth has come in closer to 8 to 15% as the broader software demand environment normalized. Customer acquisition costs have risen. Net retention has compressed. The companies are not failing. They are not even underperforming dramatically against operational benchmarks. They are simply worth substantially less than the acquisition price, and the gap is not closing.
Thoma Bravo's reported handover of Medallia to its creditors in mid-2026 is the most visible example of this pattern. A $6.4 billion 2021 take-private deal that could not service its capital structure as growth normalized and rates rose. The equity stake gets written off. The thesis gets archived. The fund vintage takes a markdown. But for every Medallia that becomes a headline, there are dozens of portfolio companies in similar situations that are being held quietly at marks that imply eventual exit value the market is unlikely to confirm.
The accumulation across the industry is significant. Industry estimates suggest 32,000 PE-backed companies are currently sitting on sponsor balance sheets without realistic exit paths at current marks. The aggregate unrealized value sits in the trillions of dollars. The runoff of that inventory will dominate PE return distributions for the next five years.
LPs are aware. The implication for new fund commitments is that LP underwriting now applies a meaningful discount to claimed TVPI from recent vintages. A 1.8x reported TVPI from a 2019 vintage gets read as something closer to 1.3 to 1.5x in LP committee discussions. The discount varies by GP, by sector, and by quality of marking history. But the practice of taking reported TVPI at face value has ended.
For the broader context on how LPs are evaluating fund managers in this environment, see what LPs actually want in 2026.
What This Does to Fundraising
GPs raising new funds against this backdrop face a structural problem. The track record is what it is. Recent vintage DPI is below historical norms across nearly every manager. The question is how to position around that reality rather than against it.
Five positioning moves have worked in current LP conversations.
Lead with realized DPI from earlier vintages. If the firm has a 2014 or 2015 fund that delivered 1.5x+ DPI, that becomes the lead artifact rather than the most recent fund's TVPI. The narrative becomes: this is what our investment strategy actually produces when given a normal exit environment. The recent vintages are working through unusual market conditions but reflect the same underlying investment discipline.
Separate the unrealized from the realized in every chart. LPs are skeptical of TVPI in 2026. GPs that show a clear breakdown of realized DPI versus unrealized RVPI in their pitch materials, with specific portfolio company detail on the unrealized portion, earn credibility points. The opacity of blended TVPI numbers has become a flag.
Document the exit pipeline explicitly. LPs want to see which portfolio companies have active sale processes, which are in management dialogue, which have refinancing planned, and which are being held for further value creation. A pipeline that converts assumption-based exit timing into concrete process detail moves the conversation from "trust us on the marks" to "here is the path to realized cash."
Address the continuation vehicle question directly. Most established GPs now have at least one continuation vehicle in their history or pipeline. LPs are not opposed to these structures, but they want to understand the economics. Who underwrites the price? What were the alternatives considered? How was LP-side liquidity handled for those who did not roll? GPs that treat continuation vehicles as a normal tool with clear governance answer this confidently. GPs that avoid the question raise more concerns.
Build distribution scenarios into Fund N+1 modeling. Forward modeling that explicitly addresses how the new fund will avoid the distribution profile of the prior fund matters. Concrete commitments on hold period discipline, leverage structure, and exit timing. LPs are not satisfied with "we will be different this time" framing. They want to see structural changes in approach.
For more on how this connects to the broader LP allocation environment, see how long it takes to raise a fund in 2026.
What Will Eventually Reset the Math
The current distribution drought will not last forever. The historical pattern across PE cycles is that liquidity restores in waves. Three forces are likely to drive the reset.
Public market multiple recovery is the first. If software, healthcare, and industrial multiples expand meaningfully, the auditable fair value of unrealized portfolio assets rises. Strategic and IPO exit windows reopen. The mark-to-cash gap compresses. The 2009 through 2013 recovery followed this pattern, with public multiple expansion driving substantial exit value across vintages that had been stuck since 2008.
Rate normalization is the second. If base rates decline from current levels, the capital structures sitting on recent vintage portfolio companies become serviceable, refinancing windows open, and dividend recaps become viable again at lower costs. Even a 100 to 150 basis point decline in rates would materially change exit economics for a meaningful portion of the existing portfolio inventory.
Forced clearing is the third. Eight to ten-year fund lives create a hard exit constraint. GPs cannot hold portfolio companies indefinitely. As funds approach end of life, sales happen at clearing prices, even when those prices imply markdowns from carrying value. The cohort of 2014 through 2018 vintage funds reaching final wind-down between 2026 and 2030 will produce a substantial volume of forced exits. The clearing process is painful for current holders but resets the comparable transaction data that LPs use to underwrite future vintages.
For LPs, the strategic question is not whether to remain in the asset class. It is which managers in the existing portfolio will exit the drought with credibility intact and which will be permanently impaired. The cohort returns from 2017 through 2021 will sort that out over the next 36 to 48 months. The managers who exited their portfolios responsibly, who marked their books conservatively, and who communicated honestly with LPs through the drought will be in a structurally stronger position to raise successor funds. The ones who relied on aggressive marks, aggressive continuation vehicles, and NAV loans to manage optics will find that the optics no longer hold.