THINGS THAT ARE FINE
MEMO: THINGS THAT ARE FINE
TO: Everyone who said private credit was mature, resilient, and built for volatility
FROM: Reality
DATE: March 13, 2026
RE: It is not fine
Let's begin with a thought experiment.
You build a $2 trillion asset class over fifteen years. You market it as the sophisticated alternative to public markets — higher yield, lower volatility, institutional-grade underwriting. You recruit retail investors into the tent by wrapping the funds in quarterly redemption windows. You tell them it's illiquid in principle, liquid enough in practice. You collect the fees. You grow the AUM. You give speeches at conferences about how private credit has matured into a genuine pillar of the financial system.
Then a war starts, the economy prints 1.4% GDP growth, core PCE prints 3.0%, software valuations start to crack, and suddenly those retail investors — the risk-averse ones who never belonged in the fund in the first place — would like their money back.
What do you do?
You gate them.
This week, Morgan Stanley's North Haven Private Income Fund told investors it would honor 45.8% of redemption requests. Investors had sought to pull nearly 11% of shares outstanding — more than double the fund's 5% quarterly cap. The fund returned roughly $169 million. The rest of the request sits in a queue, politely waiting for better days. Cliffwater's $33 billion flagship vehicle fared similarly, limiting redemptions to 7% after investors tried to pull a record 14% in Q1.
MS shares fell 4.1% on the news. Which is the market's way of saying: we noticed.
But here's the part that gets buried in the earnings footnotes. Blue Owl halted redemptions at a key fund last month. BlackRock restricted withdrawals from the HPS Corporate Lending Fund earlier in March. Blackstone disclosed its private credit fund faced a surge in Q1 redemption requests. JPMorgan has been marking down the value of certain loans, mainly in the software sector. Every major name in the space, within weeks of each other. A synchronized event, dressed up as a series of isolated operational decisions.
The phrase for this, when it happens in banking, is "loss of confidence." When it happens in shadow banking, apparently the phrase is "managing liquidity in line with fund documents."
The deeper story here is the AI unwind eating private credit from the inside.
Eroding faith in software companies was a key factor in the recent wave of redemptions, amid growing concerns that AI could dent their businesses and in turn affect their ability to repay loans. North Haven alone spans 312 borrowers across 44 industries — most of them middle-market U.S. firms that took on variable-rate debt during the era of cheap money and are now servicing that debt in a 3.5%-3.75% rate environment, while watching their software subscriptions get disrupted by the same technology wave that was supposed to be everyone's ticket to growth.
JPMorgan has started restricting lending to loans associated with software companies in its private credit funds as a precautionary measure. When JPMorgan tightens its grip on something, it is not a precaution. It is a diagnosis.
The optimists will tell you that redemption caps are written into fund documents for exactly this reason. That they are "stabilizers." That you cannot create liquidity from an illiquid asset class. All technically true. Also: the same logic applies to every gating event in financial history before it became a crisis. The gates were always "in the documents." The documents were always "protecting long-term investor value." The long-term investors were always, eventually, fine.
The short-term ones were not.
Now overlay the macro.
Q4 2025 GDP came in at a meager 1.4%, significantly hamstrung by the 43-day government shutdown, while core PCE printed 3.0% annually — above the Fed's target and above consensus. This morning, January PCE lands, and the number is already irrelevant before it hits the wire — because January was before the Strait of Hormuz became a war zone and before Brent crude crossed $100. Whatever the January data says, the February and March data will be worse. Probably significantly worse.
The Fed's problem is geometric, not arithmetic. Rates at 3.5%-3.75%, GDP at 1.4%, core inflation at 3.0%, energy prices surging, and a private credit market with $2 trillion in assets showing its first serious signs of internal stress. Every variable is moving in a direction that makes the other variables harder to manage.
The KRE Regional Banking ETF dropped nearly 5% after the PCE data landed, as fears of a renewed credit crunch resurfaced. Regional banks are being squeezed from both ends: paying higher rates to retain deposits while loan demand stalls in a slowing economy.
Goldman revised its U.S. PCE forecast upward and cut its GDP outlook. Translation: they now see stagflation as the base case, not a tail risk.
Back to the memo format, because this deserves to be said plainly.
The private credit market spent a decade telling institutional investors — and, more recently, retail investors — that it had solved the liquidity-return tradeoff. Higher yield, managed risk, periodic windows for redemption. The pitch was coherent during a decade of zero rates and benign credit cycles. It looks considerably less coherent when GDP is 1.4%, software borrowers are getting repriced for AI disruption, oil is above $100, and the largest funds in the space are returning investors less than half of what they asked to withdraw.
Default rates are approaching 9% for some funds. The baseline optimistic scenario, where the Fed cuts rates and eases the pressure on variable-rate borrowers, keeps getting pushed further out on the calendar.
No one is calling this a crisis. They're calling it a "period of heightened scrutiny." Which is, historically, the sentence that appears in documents about six months before someone starts calling it a crisis.
There is one silver lining, genuinely.
Saudi Arabia's ETF (KSA) is the only major equity market index in the green for March, up 2.3%. Riyadh is having a very good war. The rest of us are managing our redemption windows.