Four Banks Under Review: Why Private-Credit Collateral, Not Bank Losses, Is the First Signal
The New York Fed's review is not evidence of bank losses. Its real signal is that private-credit collateral marks could squeeze borrowing capacity before defaults appear.
The Federal Reserve Bank of New York has been asking four major banks a question that sounds technical but reaches into the plumbing of private credit: what, exactly, backs the money banks lend to private-credit funds? The review is not evidence that those banks have suffered losses, and it is not a public diagnosis of a systemic crisis. Its sharper signal is about the collateral marks that determine how much borrowing capacity a fund can retain when its own loans come under pressure.
Since spring, New York Fed officials have visited JPMorgan Chase, Wells Fargo, Barclays and Morgan Stanley to ask about their exposure to private-credit firms, risk controls and collateral quality, according to Semafor's Oct. 5 report. The review was prompted in part by JPMorgan's March decision to mark down loans in private-credit portfolios, particularly exposures to software companies. Some bank reviews, including JPMorgan's, have concluded, Semafor reported, while the Fed and the banks declined to comment.
That sequence matters. A supervisory visit is a request to understand a risk, not proof that the risk has crystallized. But the collateral question is more consequential than a headline tally of loans: private-credit funds often borrow from banks against portfolios of loans they have made to companies. If those underlying loans are marked lower, a bank can reduce the amount it is willing to advance against them. A fund may then have to post more collateral, repay some borrowing or seek liquidity elsewhere, even if the borrower has not missed a payment.
This is why the first transmission channel may be a tighter borrowing base rather than a sudden wave of bank charge-offs. A haircut is not the same thing as a default, and a more conservative valuation is not automatically a fire sale. Yet if several lenders reassess similar collateral at the same time, funding can become less available precisely when a fund's investors are asking for liquidity or the fund is trying to support portfolio companies. In private markets, where there is no continuous exchange price for every loan, the timing and consistency of those marks can shape the speed of adjustment.
The scale deserves careful description. The FDIC's 2026 Risk Review put bank loans to nondepository financial institutions, or NDFIs, at $1.4 trillion as of Dec. 31, 2025, equal to 5.6% of industry assets. That is a broad category, not a private-credit-only number: it includes mortgage and business credit intermediaries, insurers, private equity funds, securitization vehicles and other entities. The FDIC also said ten banks held about 66% of NDFI loan balances, while banks with more than $100 billion in assets held 86%.
The reporting system is still catching up with the market. The FDIC said NDFI lending rose 35.2% year over year in 2025, while warning that some of the increase reflected banks reclassifying loans after reporting instructions changed. This caveat cuts both ways. The broader connections are large enough to merit attention, but the aggregate does not tell an investor how much is private-credit lending, how much is secured by a fund's portfolio, or which positions are exposed to the same underlying borrower or sector.
Nor do current bank-loan performance data read like an immediate credit event. The FDIC reported that the past-due and nonaccrual rate for NDFI loans was 0.15% in the fourth quarter of 2025, compared with 1.39% for commercial and industrial loans. It described most NDFI facilities as short-term revolving credit lines, generally supported by conservative advance rates. Those numbers are reassuring about reported performance; they cannot by themselves settle whether collateral values are current, whether funds can meet redemptions or how correlated loan marks might behave in a broader downturn.
The global picture is also wider than one bank review. The Financial Stability Board estimated private-credit assets at $1.5 trillion to $2 trillion at the end of 2024. Available data from FSB members captured around $220 billion in drawn and undrawn direct bank credit lines to private-credit funds, while commercial estimates ranged from $270 billion to $500 billion. Those figures are not interchangeable: the FSB highlighted inconsistent definitions and missing fund- and loan-level data, and the captured credit-line figure does not include every indirect exposure.
That distinction is central to the story. A bank line to a fund is only one link. Banks can also lend to companies that borrow from private-credit funds, provide financing to insurers or other investors in the funds, or participate in synthetic risk transfers. The FSB has described these interconnections as deepening, while emphasizing that private credit has not been tested through a prolonged, severe downturn at today's scale. The result is a measurement problem as much as a loss problem: regulators need to see whether exposures are concentrated, how collateral is valued and where liquidity demands could travel.
The software loans that helped trigger the New York Fed's review illustrate the difficulty. An AI-driven change in the outlook for a software company can alter lender assumptions about future cash flow, but a private loan may not trade often enough to produce a clean market price. If the borrower remains current, an unchanged mark may look defensible to one lender and stale to another. If the mark is cut, the question for a bank is not simply whether the loan will default; it is whether the haircut, advance rate and documentation still protect the facility it extended to the fund.
The regulatory signal has sharpened. On Sept. 28, SEC Chief Accountant Kurt Hohl and Brian Daly, director of the Division of Investment Management, issued a staff statement on private-asset valuations. They described private-credit assets as “typically illiquid, individually negotiated loans that do not trade on established secondary markets and therefore generally lack readily available quoted prices.” The statement did not create new rules. It reminded registrants that fair-value estimates should reflect a market participant's perspective, use robust calibration and disclose material assumptions clearly.
That focus on marks was visible before the latest review. Michael Barr, a Federal Reserve governor, told Bloomberg News in a May 3 interview that direct links between banks and private credit did not yet appear “super worrisome,” while warning that trouble could provoke “psychological contagion.” His distinction is useful: a narrow bank exposure may be manageable in isolation, but anxiety can broaden if investors decide that weak marks in one corner are evidence of a wider credit problem.
Jamie Dimon, JPMorgan Chase's chief executive, gave the more contained view in remarks reported by Reuters on April 14. Asked by analysts whether private-credit risks were systemic, he said: “I don't think it's systemic.” That is not a claim that every fund is safe or that every bank line is insulated. It is a judgment about the size and shape of the direct channel, and it sits alongside his firm's decision to reassess collateral values.
The tension between those views is more informative than either slogan. A sector can fall short of a system-wide solvency threat and still produce a painful liquidity cycle for funds, lenders and middle-market companies. A smaller pool of bank credit can raise the price of refinancing or force borrowers to seek alternatives. Meanwhile, private-credit investors may be unable to exit at a stable published price, and managers may be limited in how quickly they can convert loans into cash. The pressure can therefore show up in lending terms and redemptions well before it appears as a wave of bank defaults.
Our view is to treat the New York Fed's review as a collateral and funding signal, not a verdict on private-credit solvency. Investors should separate the $1.4 trillion of broad NDFI lending from direct private-credit exposures, and distinguish a lower borrowing base from a realized loss. The practical watch list is the next round of bank disclosures and fund reports: changes in advance rates, borrowing-base availability, valuation frequency, software-loan marks, redemption requests and any evidence that several lenders are applying the same haircut at once.
The better early-warning indicator may be the gap between reported loan performance and the amount of financing a fund can still draw. If that gap widens, the result could be a selective credit squeeze without a banking crisis. If marks hold up, liquidity remains available and default rates stay contained, today's supervisory attention may prove to be routine prudence. For now, the Fed's questions point to where the market's next argument will be fought: not whether private credit exists on bank balance sheets, but whether its collateral can keep supporting the same amount of credit when lenders become less certain about its value.
This material is for informational purposes only and is not investment advice or a recommendation to buy or sell any security or financial instrument. The views expressed are subject to change without notice. Readers should conduct their own research and consult qualified advisers before making investment decisions.
