Omega Ukama — Financial Journalist & Researcher | About the author →
Yale was finalizing a sale of up to $2.5 billion of private-equity and venture-capital interests in June 2025. Harvard was in advanced talks over a roughly $1 billion sale. Neither transaction proved that the endowment model had failed, and neither was reported as a fire sale.
They were reminders that a portfolio can be solvent, sophisticated and built for decades - and still need to turn illiquid assets into cash at a time it did not choose.
That distinction matters for every large owner with a private-market programme. The illiquidity premium is not an accounting fiction. Patient capital can earn returns unavailable to an investor forced to sell on demand. But the premium belongs only to an institution able to carry the asset through the full period in which the cash is unavailable.
The danger is not one hard-to-sell fund. It is several claims on the same liquid pool arriving together.
Four bills can land on one calendar
1. Capital calls become due
Private-equity, infrastructure and private-credit commitments are legal obligations. Notice is often measured in days, not months. Pacing models can estimate calls by vintage and strategy, but they cannot guarantee that managers will slow deployment when markets fall. Some funds call capital during stress to finance acquisitions, protect portfolio companies or support follow-on investments.
2. Benefits and operating draws continue
Pension payments, insurance claims and endowment spending do not wait for a better exit market. A mature pension plan may already pay more in benefits than it receives in contributions. A university may depend on its endowment for a material share of annual operations. These demands may be predictable, but they are not optional.
3. Collateral is demanded immediately
Currency and interest-rate hedges can require cash or government bonds when markets move. The 2022 British liability-driven investment crisis demonstrated the feedback loop: higher gilt yields produced collateral calls; pension funds sold gilts to raise cash; those sales pushed yields higher and created further calls. Private assets did not cause the shock. They reduced the portion of the portfolio available to meet it.
4. Expected liquidity is prorated
Closed-end institutional funds generally do not promise redemptions. Evergreen, interval and non-traded vehicles have changed the boundary by offering periodic repurchases, normally subject to caps. In the second quarter of 2026, investors sought to redeem 10% of shares at Blackstone's BCRED, 17% at Cliffwater's corporate lending fund and 16.8% at Apollo Debt Solutions. Each paid about 5% under its disclosed limit.
The companion article in this package examines those mechanics. The portfolio-level point is simpler: the amount not paid becomes the allocator's cash problem.
Where the standard model can break
Most liquidity frameworks are sensible. They project capital calls and distributions, maintain cash and public assets, and stress-test market declines. The weakness appears when the model assumes that each demand channel will retain the same relationship to markets that it had during years of abundant distributions and functioning dealer balance sheets.
A public-market decline can push all four channels at once. It shrinks the liquid assets serving as the buffer. It may increase collateral requirements. It can slow private-market exits and distributions. It may prompt managers to call capital for defensive financings. It can also make investors in semi-liquid vehicles request cash while private valuations adjust only slowly.
The denominator effect makes the problem look smaller before it looks larger. Public holdings reprice immediately; private holdings are valued less frequently and usually more smoothly. As public assets fall, private assets become a larger reported share of the portfolio even without a new commitment. An owner can breach its allocation ceiling at the same moment its practical ability to rebalance is weakest.
A liquidity buffer is therefore not a fixed percentage. It is a set of assets whose value, market depth and collateral eligibility can deteriorate at the same time the institution needs them.
The exposure differs by owner
Endowments combine large private allocations with annual operating draws and unfunded commitments. Yale's reported sale should be read as active portfolio management under a changed distribution environment, not evidence of insolvency. The important questions are the price paid for liquidity and the future returns surrendered.
Public pension funds face a lifecycle problem. A growing plan with positive net contributions has an internal source of cash. A mature plan paying more in benefits than it receives has less room for error. The same private allocation can carry very different risk depending on the plan's cash-flow status, derivative programme and access to committed funding.
Insurers face both cash and regulatory-capital channels. A rating or valuation change can increase capital pressure while claims or collateral calls consume liquidity.
Sovereign wealth funds often have fewer contractual liabilities, but a government may call on a fund during the same shock that depresses portfolio values or commodity revenue. The reserve can dry up just as the sponsor needs it most.
Family offices may have less formal liquidity infrastructure and a greater chance that portfolio demands collide with operating-business or family cash needs.
The retirement channel raises a new design question
In March, the US Department of Labor proposed a safe harbour for defined-contribution fiduciaries selecting professionally managed funds containing alternative assets. US 401(k) plans held about $9.9 trillion at the end of the first quarter.
Even a modest private-market allocation would bring a different pattern of withdrawals: retirements, hardship needs, loans and participant reallocations occur continuously, not according to an institutional pacing calendar. The policy case for wider access is understandable. Workers should not be excluded automatically from assets offered to wealthy and institutional investors. But a daily-valued retirement plan and a portfolio of multi-year, non-traded loans do not naturally fit. The wrapper, liquidity sleeve and valuation process may matter as much as the allocation.
What boards should require
A portfolio-wide liquidity test should model simultaneous, not sequential, demands.
- Model cash needs by legal due date, not annual average.
- Separate contractual liquidity from assumed market liquidity.
- Reduce reported private values and expected secondary-sale proceeds under stress.
- Stress the same borrower across private debt, public equity, insurance and bank exposures.
- Test repo, committed credit lines and eligible collateral under a market-wide shock, including counterparty limits.
- Pre-authorize who may sell assets, draw a line, slow commitments or exceed an allocation band.
- Run the test through several quarters, because prorated redemptions and weak distributions can persist.
What to watch next
The most useful indicators are observable: repurchase requests relative to each fund's cap; the difference between requested and paid amounts; distributions as a share of private-market value; unfunded commitments relative to genuinely liquid assets; collateral usage; secondary-market discounts; and the share of benefits or operating costs funded from current contributions or revenue.
Private markets did not mislead investors by being illiquid. The governance failure is treating illiquidity as a static allocation percentage rather than a claim on future cash. The premium remains valuable. The buffer is credible only if it survives the quarter in which all four bills arrive together.
Sources and methodology
Reuters: Yale was finalizing a sale of up to $2.5 billion of private-equity interests
Reuters: Harvard was in advanced talks over a roughly $1 billion secondary sale
Reuters: BCRED received second-quarter requests equal to 10% of shares
Reuters: Cliffwater received requests equal to 17% of shares
Reuters: Apollo received requests equal to 16.8% of shares
Financial Stability Board: vulnerabilities in private credit
Federal Reserve: Financial Stability Report, May 2026
European Central Bank: Financial Stability Review analysis of private credit
US Department of Labor: proposed alternative-assets safe harbour
Investment Company Institute: US retirement assets, first quarter 2026
The reported Yale and Harvard processes are not described as completed transactions. Fund-request figures use the same second-quarter definitions as the companion private-credit article.