Yale did not have a solvency problem. That is why the signal matters.
In April 2025, Yale confirmed it was exploring a sale of private-equity fund interests and had retained Evercore as an adviser. By June, Reuters reported that the university was nearing completion of a sale of up to $2.5 billion of private-equity and venture-capital assets in a transaction code-named “Project Gatsby.” Harvard was moving along a similar track: Reuters reported that Harvard was exploring a roughly $1 billion sale of private-equity fund interests amid financial uncertainty and political pressure.
Neither institution was a distressed seller in the conventional sense. Yale later reported an 11.1% investment return for fiscal 2025, representing $4.5 billion of investment gains. After distributing $2.1 billion to the operating budget, Yale’s endowment rose to $44.1 billion as of June 30, 2025.
That is the point. The issue was not bankruptcy. It was plumbing.
For the world’s largest long-horizon investors, the old question was: Can we tolerate illiquidity?
The better question now is: Can we fund illiquidity when every demand for cash arrives on the same calendar?
That is the liquidity illusion. Private markets promised a premium for investors who did not need to sell. They did not promise that capital calls, operating draws, benefit payments, collateral calls, redemption requests, and sponsor needs would remain politely uncorrelated forever.
| Download the deck (PDF) → |
The source trail is the story
The warning does not appear in one place.
It appears in fragments.
In university endowment transactions, where Yale and Harvard moved toward secondary sales even though neither institution was facing a conventional solvency crisis. In private-credit fund letters, where investors requested more liquidity than semi-liquid vehicles were designed to provide. In European pension stress tests, where margin calls turned market moves into immediate cash needs. In Financial Stability Board work, where regulators now describe private credit through the language of interconnectedness, leverage, liquidity mismatch, opacity, and data gaps. In secondaries-market reviews, where record transaction volume shows that liquidity exists, but only at a clearing price. And in private-equity cash-flow modeling research, where the old Takahashi-Alexander framework is being reworked because private-market cash flows are not only a function of fund age; they are also a function of macro conditions and GP discretion.
No single source says the universal-owner liquidity model is breaking.
That is the point.
The risk is visible only when the sources are read together.
The premium was real. The assumption underneath it was fragile.
Private markets earned their place in institutional portfolios for structural reasons.
A pension fund, sovereign wealth fund, insurer, endowment, or large family office has one advantage short-term capital does not: patience. It can accept lockups. It can ride through public-market volatility. It can underwrite complexity. It can commit capital when others need liquidity.
That advantage is still real. But it is not free.
The private-market premium is compensation for giving up an option: the option to sell quickly, cleanly, and near reported value. For much of the post-2008 era, that option appeared cheap. Rates were low. Exits were open. Distributions were strong. Public markets kept rising. Private marks moved slowly enough to make volatility look manageable.
That regime has changed.
McKinsey’s 2026 private-equity work argues that the conditions that once amplified returns — declining rates, multiple expansion, and abundant leverage — have passed. In 2025, buyout funds underperformed U.S. and global public equities for the third consecutive year, generating roughly 7% returns compared with 18% for the S&P 500 and 22% for MSCI World. McKinsey also reported that distributions as a percentage of AUM fell to about 6% in the six months ending June 2025, compared with a ten-year average of 14%.
The conclusion is not that private markets no longer work. It is that the premium is conditional. It depends on manager selection, entry price, vintage, leverage, fees, exit conditions, governance, and the owner’s ability to remain a liquidity provider when others become liquidity takers.
That last condition is now the one boards should test hardest.
A fund can be long horizon and still need cash next Thursday. A university cannot defer payroll because distributions slowed. A mature pension plan cannot suspend benefits because capital calls accelerated. An insurer cannot ignore collateral because private assets are marked quarterly. A sovereign fund cannot tell its sponsor government that the portfolio’s horizon is intergenerational when the treasury’s shortfall is this fiscal year.
Patience is an investment advantage. It is not a cash-management plan.
Four bills, one calendar
The first bill is the capital call.
A private-fund commitment is not a soft allocation. It is a contractual obligation. When a general partner calls capital, the limited partner wires money. A pacing model can estimate timing, but it cannot veto the notice. An owner that committed $100 million and still has $60 million unfunded does not get to renegotiate the exposure because public markets are down.
The second bill is the operating draw.
For an endowment, that is institutional spending. For a pension fund, it is benefit payments. For a sovereign fund, it may be fiscal transfers. These obligations do not disappear in a drawdown. In some cases, they become more politically urgent precisely when markets, tax receipts, or sponsor budgets are under pressure.
The third bill is collateral.
Large asset owners routinely use derivatives to hedge interest-rate, currency, inflation, or liability risk. Those hedges may be prudent, but they require cash or government securities when markets move against the position. The UK’s 2022 liability-driven investment crisis showed the danger: rate moves and margin calls can force selling into stressed markets, turning a hedge into a liquidity amplifier.
EIOPA’s 2025 occupational pension stress test made the same point in regulatory language. The exercise covered 156 institutions for occupational retirement provision across 18 countries, representing about 60% of the European Economic Area IORP market by assets. EIOPA concluded that the sector had sufficient liquidity buffers in aggregate, but that margin calls can still pose a threat, particularly for derivatives users.
The fourth bill is redemption pressure.
Most institutional private-market exposure sits in closed-end vehicles. But the fastest-growing edge of private markets is no longer purely institutional or fully locked. Semi-liquid private-credit funds, interval funds, non-traded BDCs, and wealth-channel structures offer periodic liquidity, usually capped at a percentage of NAV.
In calm markets, that looks like convenience. In stress, it becomes a queue.
Blue Owl’s 2026 experience shows the mechanism. Reuters reported that investors requested $4.7 billion of withdrawals from two Blue Owl private-credit funds in the second quarter of 2026, down from $5.4 billion in the first quarter but still elevated enough that the firm maintained 5% quarterly withdrawal caps.
Blackstone’s BCRED faced its own version of the pressure. Reuters reported in March 2026 that redemption requests totaled 7.9% of the $82 billion fund, before Blackstone allowed investors to pull more than the usual quarterly limit. In June, Reuters reported that Blackstone capped withdrawals from the flagship private-credit fund after another round of elevated redemption demand.
Cliffwater’s flagship private-credit fund was also tested. Reuters reported that investors in a $31.3 billion Cliffwater private-credit fund requested withdrawals equal to 17% of shares in the second quarter of 2026, while redemptions were capped at 5%.
The lesson is not that these vehicles are broken. Many are operating exactly as designed.
The lesson is that the design matters.
A gate is not a failure of the contract. It is the contract reminding investors what they bought.
The hidden risk is correlation
Each liquidity demand is manageable in isolation.
Capital calls are manageable. Spending draws are manageable. Benefit payments are manageable. Collateral calls are manageable. Redemption queues are manageable. Fiscal-transfer requests are manageable.
The risk is simultaneity.
A public-market drawdown is often the event that makes the other channels move together. Falling equities shrink the liquid side of the portfolio. Wider spreads slow exits. Slower exits reduce distributions. Lower distributions make new commitments harder to fund. Rate or currency moves trigger collateral calls. Semi-liquid investors submit redemption requests. Sponsor governments or universities demand cash at the same time the portfolio has less free liquidity.
That is where the standard model breaks. Most liquidity frameworks still treat liquidity as a reserve. Boards should increasingly treat liquidity as a correlation problem.
Regulators are moving in that direction. In May 2026, the Financial Stability Board warned that private credit’s complexity, leverage, and interconnectedness could amplify stress in adverse scenarios. The FSB also highlighted liquidity mismatches, insurer links, multiple layers of leverage, valuation opacity, and data gaps.
The issue is not whether a single asset class fails in isolation. It is whether several balance-sheet claims become cash demands at once.
The denominator effect is only the visible part
Most boards already understand the denominator effect. Public markets fall. Private marks lag. Private assets become a larger share of the total portfolio. The owner appears overallocated precisely when liquidity is scarcer.
S&P Global Market Intelligence reported that about 62% of global pension funds exceeded their private-equity allocation targets at the end of June 2025. The median actual allocation across 294 pension funds was above the median target allocation.
That is usually framed as an allocation problem. It is actually a governance problem.
An owner that moves from 18% private equity to 23% private equity because public markets fell has four bad choices. It can stop making new commitments and risk losing manager access. It can sell secondaries and crystallize a discount. It can rebalance by selling liquid assets, further reducing the liquidity buffer. Or it can tolerate the overshoot and accept less flexibility.
Meanwhile, distributions are not recycling capital the way LPs became accustomed to in the prior regime. McKinsey found that LP priorities have shifted decisively toward liquidity, with delayed exits and lack of liquidity ranking among the top LP concerns.
Low distributions are not only a return problem. They are a liquidity problem. They reduce the internal cash recycling that allows owners to fund new commitments without selling something else.
The private-equity industry still has enormous dry powder, but the number should be understood from both sides of the table. S&P Global Market Intelligence, using Preqin data, put global private-equity dry powder at $2.184 trillion as of March 31, 2025, down from a record $2.305 trillion at year-end 2023.
For GPs, dry powder is capital waiting to be deployed.
For LPs, it is also a future invoice.
The model risk hiding inside the liquidity risk
Most investment committees do not manage private-market liquidity by instinct alone. They use pacing models. They forecast capital calls, distributions, NAV growth, and unfunded commitments. They build liquidity ladders. They ask how much cash needs to be available under base, downside, and stress scenarios.
The problem is that many models still treat private-market cash flows as if they are mostly a function of fund age.
That is only partly true.
The industry workhorse is the Takahashi-Alexander model, sometimes associated with the Yale model. It became popular because it is simple, intuitive, and practical. A fund calls capital early, invests it, grows NAV, and distributes capital later. The shape is the J-curve. The model gives LPs a way to forecast contributions, distributions, and residual value using assumptions such as call rate, growth rate, fund life, and the “bow factor” that shapes the timing of distributions.
That simplicity is also the weakness.
A 2024 Stockholm School of Economics master’s thesis by Gustav Erikson and Måns Holmberg revisited the Takahashi-Alexander model using Burgiss data covering 2,671 global buyout funds with vintages from 1990 through 2022. The authors adapted the model to reflect macroeconomic effects on three parameters: the rate of capital contributions, NAV growth, and the bow factor that governs the phasing of distributions. Their conclusion is highly relevant for universal owners: macro-calibrated modeling significantly improved prediction of capital contributions, but prediction of net cash flows remained less definitive because modeling distributions proved more complex.
That is the asymmetry boards should care about.
Capital calls are more predictable than distributions. Obligations are easier to forecast than relief.
A model can say distributions should arrive in year seven. The market can say exits are closed. A model can say capital calls should slow. A GP can find a rescue financing, add-on acquisition, or opportunistic entry point. A model can say the private allocation is within range. A public-market drawdown can make it overallocated by Monday.
This is the underappreciated governance problem.
In public markets, an owner may not like the price, but it can sell. In private markets, the owner may like the valuation, but it cannot force the distribution.
The GP does not only manage the asset.
The GP manages the clock.
And in a liquidity crisis, the clock is the asset owner’s scarcest resource.
The secondaries market is the counterargument — and the warning
The private-market defense deserves to be taken seriously.
The secondaries market is no longer a distressed back alley. It is a core portfolio-management tool. Jefferies reported that global secondary-market transaction volume reached a record $240 billion in 2025, up 48% year over year. GP-led secondary volume reached $115 billion, or 48% of total secondary-market activity.
This matters. Secondaries, continuation vehicles, NAV financing, GP-led transactions, and structured liquidity solutions give LPs more options than they had a decade ago. The industry has adapted.
But adaptation is not immunity.
Secondaries do not eliminate illiquidity. They reveal its clearing price.
That distinction matters. A fund interest marked at 100 and sold at 90 did not suddenly become risky on the day of sale. The risk was already there. The sale only made it visible.
For boards, the key question is not whether liquidity exists. In most cases, it does. The question is what the institution must give up to obtain it when everyone else wants the same thing.
The exposure looks different for every owner
The liquidity illusion is not the same across institutions. Treating it as a single “private markets” issue obscures the real vulnerabilities.
Endowments face the operating-budget version. They often combine high private-market allocations with recurring institutional spending. Yale is useful as a case study not because Yale is weak, but because Yale is sophisticated. If a top-tier investment office uses the secondaries market for portfolio flexibility, boards elsewhere should treat that as a signal, not an embarrassment.
Public pension funds face the lifecycle version. A young plan receiving more contributions than it pays out can ride through a liquidity squeeze. A mature plan paying out more than it collects has less room. Derivatives can also convert market moves into immediate cash needs, as EIOPA’s stress test showed.
Insurers face the capital version. Private credit and illiquid fixed-income assets may match long liabilities well, but they also introduce opacity, ratings dependence, valuation uncertainty, and regulatory-capital questions. The FSB specifically flagged insurer interlinkages and the use of private credit ratings as issues requiring attention.
Sovereign wealth funds face the sponsor version. Many do not have pension-style benefit payments, but commodity-linked and fiscal-stabilization funds can face government drawdowns precisely when oil, equities, or domestic revenue are under pressure. State Street reported that sovereign wealth fund investments in private markets reached $3.5 trillion in its sample by the end of 2025. CFA Institute, citing State Street, noted that average private-market exposure rose from around 25% in 2020 to nearly 30% by the end of 2025.
Family offices face the governance version. Many have sophisticated portfolios but thinner liquidity infrastructure than large pensions or insurers. Their risk is not only a capital call. It is a capital call landing beside a tax event, operating-company need, real-estate refinancing, divorce settlement, philanthropy pledge, or family liquidity request. Institutional capital usually has committees. Family capital often has surprises.
The same asset can therefore be safe for one owner and dangerous for another.
Illiquidity is not an asset-class label. It is an interaction between the asset, the liability, the governance system, and the calendar.
The fifth liquidity channel: retirement savers
The next liquidity channel may come from outside the institutional portfolio.
In March 2026, the U.S. Department of Labor proposed a rule clarifying fiduciary duties and providing a safe harbor for the selection of designated investment alternatives in participant-directed retirement plans. The proposal was part of the policy push to expand access to alternative assets inside 401(k) plans and other defined-contribution structures.
The market is enormous. Americans held $13.8 trillion in employer-based defined-contribution retirement plans on March 31, 2026, including $9.9 trillion in 401(k) plans, according to the Investment Company Institute.
For private-market managers, that is a distribution opportunity.
For universal owners, it is also a market-structure issue.
Defined-contribution plans have a different liquidity profile from institutional separate accounts. Individual savers retire, rebalance, switch plans, take hardship withdrawals, borrow, panic, and follow target-date flows. Those demands are granular, continuous, and behaviorally sensitive.
That does not mean private assets are automatically inappropriate for DC savers. It means vehicle design matters more than marketing.
A private asset can be suitable for a long-horizon saver and still be dangerous inside a structure that promises liquidity it cannot reliably deliver in stress.
This is where the institutional and retail worlds are starting to collide. The same asset managers that serve sovereign funds, pensions, insurers, and endowments may also be asked to manage daily participant behavior inside semi-liquid structures.
That creates a new demand channel in the same private-market ecosystem institutional owners already depend on.
The retailization of private markets is not just a democratization story. It is a liquidity-transmission story.
Sources to watch
For boards and CIOs, the most useful early-warning indicators are not headlines. They are source trails.
Watch private-credit fund shareholder letters for redemption requests versus quarterly caps.
Watch secondaries pricing reports for the discount between reported NAV and executable liquidity.
Watch pension stress-test results for margin-call resilience after management actions, not before them.
Watch private-equity distribution yield as a percentage of NAV, because low distributions are not just a return problem; they are a cash-recycling problem.
Watch unfunded commitments by vintage year, especially funds in the first five years of life.
Watch Department of Labor rulemaking on 401(k) alternatives, because defined-contribution adoption could introduce a new behaviorally sensitive liquidity channel into private markets.
And watch the models. Any liquidity forecast that treats capital calls, distributions, collateral, and public-market drawdowns as separate events is probably measuring the wrong world.
The board dashboard needs to change
The conventional board question is: Are we within target allocation?
The better questions are harder.
How much deployable liquidity do we have after collateral needs, not before them?
What percentage of our private book is unfunded commitment rather than funded NAV?
What happens if distributions stay at half their historical rate for three years?
Which liquidity sources depend on repo markets, money-market funds, or secondary buyers functioning normally?
What private assets would we sell first, and at what discount would the investment committee stop calling it liquidity and start calling it impairment?
How much of our private-market exposure is in vehicles that can gate us, and how much of our obligation is in vehicles that can call capital from us?
That final distinction is the most important.
Redemptions can be gated. Capital calls cannot.
For a universal owner, the asymmetry matters. The cash you expect to receive may be delayed. The cash you owe may still be due.
Boards should monitor seven indicators.
First, unfunded commitments divided by deployable liquidity after collateral stress. The denominator should not be total portfolio value. It should be cash and liquid assets after margin demands.
Second, young-vintage exposure. Funds in years one to five are structurally more call-heavy. Pacing models should treat young-vintage commitments as a different liquidity risk from mature NAV.
Third, expected distributions under a closed-exit scenario. Boards should run a three-year low-DPI case, not only a return drawdown.
Fourth, recallable-distribution exposure. Treat recallable capital as a contingent liquidity claim, not an accounting footnote.
Fifth, secondaries discount to carrying value. This is the market price of liquidity, not merely a trading statistic.
Sixth, redemption requests versus gates and caps in semi-liquid vehicles. Occasional spikes are one thing. Persistent oversubscription to caps is a structural warning.
Seventh, macro-calibrated call-rate sensitivity. Pacing models should incorporate financing conditions, yield spreads, exit markets, and public-market drawdowns — not only vintage age.
The real conclusion
Private markets did not lie to institutional investors.
The illiquidity premium exists for some owners, in some vintages, with some managers, under some governance systems. Long-horizon capital still has a structural edge. The largest asset owners should not abandon private markets because a few semi-liquid vehicles showed stress.
But boards should stop talking about illiquidity as if it were only an allocation bucket.
Illiquidity is a balance-sheet design choice. It is a legal obligation, a governance constraint, a collateral demand, a valuation convention, a secondary-market dependency, and increasingly a distribution-channel risk.
Yale and Harvard did not prove the endowment model is dead. They showed that even the most sophisticated owners can discover, late in the cycle, that liquidity is not measured by what a portfolio is worth. It is measured by what the institution can fund when several claims arrive at once.
The question for every universal owner is no longer whether it can tolerate illiquidity most of the time.
It can.
The question is whether its liquidity budget was built for the world where capital calls, spending draws, benefit payments, collateral calls, sponsor demands, and redemption stress remain separate events — or for the world where they become the same event.
Only one of those worlds has been tested at today’s scale of private-market allocation.
It is not the second one.
Omega Ukama writes for the Universal Asset Owners Newsroom on private markets, institutional liquidity, and the balance-sheet design of the world’s largest asset owners. Read more on his contributor profile.