Semi-Liquid Private Credit Hits the Redemption Wall

Philadelphia, Oct. 31 — Semi-liquid private credit was built to square a circle: decade-long, unrated loans on the asset side and quarterly repurchase windows on the liability side, capped at 5% of shares outstanding each quarter. The structure worked while gross inflows exceeded redemptions and software borrowers looked stable. It is failing the stress test now that AI disruption reprices default risk, net outflows arrived, and gates left billions of requests unpaid.

Knowledge at Wharton summarized research on October 26 from finance professors Itay Goldstein and Yao Zeng (Wharton) and Chuck Fang (Drexel), drawing on 52 semi-liquid BDCs from Q1 2024 through Q2 2026. The paper—“The Fragility of Semi-Liquid Private Credit Funds”—argues redemption pressure has reached run-like proportions industry-wide, not only at a single sponsor.

The Q2 2026 ledger line

Global private credit has swelled to roughly $3.5 trillion in assets under management, but the semi-liquid slice is where liquidity promises meet loan illiquidity. Net assets in semi-liquid BDCs rose from $10 billion in 2019 to $300 billion at end-2025, according to the Wharton piece.

The breakpoint quarter is Q2 2026:

Flow (semi-liquid BDCs)	Amount
Repurchase outflows	$7.1 billion
New share issuance inflows	$3.8 billion
Net outflows (first on record)	~$3.3 billion
Redemption requests left unmet (“gated”)	$10.7 billion

The 5% quarterly cap was never meant to be a hard ceiling on investor demand—it is a rationing rule when cash, maturities, and new sales cannot fund exits. $10.7 billion of gated requests can roll forward, intensifying the next window.

Goldstein told Wharton: “If you think others are going to come and take the money out, you want to be there first.” That is textbook strategic complementarity applied to family-office and wealth-channel products marketed as semi-liquid alternatives.

Who asked out—and why software mattered

The study cites extreme repurchase demand at brand-name vehicles—including requests equivalent to roughly 22% and 40% of shares outstanding at Blue Owl Capital Corp II and Blue Owl Credit Income Corp., with Blackstone, BlackRock, and Apollo funds also breaching the 5% cap on requested redemptions.

Software exposure is the fundamental shock. Semi-liquid BDCs are “heavily exposed to software companies”—about one-fifth of loan books—with Wharton noting a “software crisis” as AI agents threaten SaaS cash flows and probability-of-default estimates jump. Markdowns on illiquid, unrated paper arrive slowly; stale net asset values let redeemers exit at prices above true economic value, shifting losses to stayers—another run amplifier.

Average portfolio facts from the research, as summarized by Wharton:

• Cash: ~5.5% of total net assets—quickly exhausted by recurring 5% quarterly redemption waves
• Net investment income: ~2.5% of TNA per quarter, largely paid as dividends (BDCs must distribute ≥90% of investment income for pass-through tax status)
• Maturing loan proceeds: <0.5% of TNA per quarter
• Unrated loans: ~80% of holdings; most names held by fewer than ten fund families—thin secondary markets

When internal cash and coupons fail, funds sell loans at fire-sale discounts, borrow from banks, or accumulate repurchases payable—economically trade credit owed to exiting investors. Wharton reports that a one-percentage-point increase in net outflows (as a share of TNA) associates with a 0.38-point rise in net debt issuance—a leverage-redemption doom loop.

Sponsors as discretionary lenders

Some managers injected affiliate capital when caps bind—Blackstone and employees adding funds to meet repurchase pressure above usual limits, per the article. Those relationship-based liquidity sources help named funds but do not scale when stress is broad-based, the authors warn.

That is the same CAPITAL hierarchy Culled tracked when Apollo bought into Mubadala’s GP and when Barclays’ trading-book booking fight moved private loan risk onto bank supervisors’ desks. Semi-liquid BDCs sit downstream of borrowers and upstream of wealth allocators who believed quarterly windows were money-good liquidity, not priority rights in a gate queue.

Retirement access raises the STATE stakes

The Wharton piece notes the U.S. Department of Labor proposed a rule in March 2026—following a presidential executive order—to encourage 401(k) and defined-contribution plans to offer alternative assets, including semi-liquid private credit. Run-like BDC stress turns that rule from a yield conversation into a fiduciary hazard: retail savers may not price gated redemptions, delayed payments, or NAV staleness the way institutional LPs do.

The Federal Reserve’s October integration of private credit into Financial Accounts data improves system-wide visibility but does not unwind vehicle-level mismatches. Culled’s earlier shadow-banking reckoning mapped covenant-lite leverage; this episode adds quarterly liquidity transformation with hard numbers on unmet requests.

What allocators should watch next quarter

Signal	Why it matters
Gated redemption backlog	Unpaid Q2 requests may stack into Q3 repurchase windows
Software markdown pace vs. repurchase NAV	Stale marks determine who subsidizes whom on exit
Net debt issuance vs. outflows	Confirms leverage doom loop in filings
Sponsor support disclosures	Discretionary capital masks—but does not remove—systemic correlation
DOL rule finalization	Extends or limits retail access to gated structures

Semi-liquid private credit sold institutional returns with a quarterly exit. Q2 2026 showed the exit is pro-rata, capped, and sometimes unpaid. Until loan books reprice fast enough to match AI-driven credit stress—or until vehicles go fully closed-end—the scarce asset in these funds is not spread; it is position in the redemption queue.