A universal owner cannot diversify away from private credit. At $1.5–2 trillion and embedded in pension, insurance and sovereign portfolios, it is now a systemic position, not a satellite one. So the question raised by the Financial Stability Board's May 6 report is not whether to be in private credit — most large owners already are, structurally — but what we need to know, and document, before this asset class meets its first severe downturn. This is a Risk Radar question, and the evidence to answer it is unusually good right now.
What the evidence says.
Four credible bodies have converged on the same shortlist of vulnerabilities, which is itself a signal worth weighting.
The FSB's Report on Vulnerabilities in Private Credit identifies leverage in opaque, multi-layered structures; liquidity risk from the rise of semi-liquid funds offering investor redemptions; concentration in technology, healthcare and services; and bank-and-insurer interlinkages the data cannot yet size — members captured roughly $220 billion of bank credit lines to private credit funds, while the regulator warns the real figure could be more than double that. The IMF reached a strikingly similar conclusion in its Global Financial Stability Report chapter, naming five vulnerabilities: fragile borrowers, semi-liquid vehicles, multiple layers of leverage, stale and subjective valuations, and unclear interconnections. The U.S. Office of Financial Research, in a March brief, put numbers on the interconnection point specifically — finding that global systemically important banks and other banks are the predominant funding source for private credit funds. And a March 2026 academic review framed the borrower-quality concern precisely: private credit borrowers are smaller, more leveraged and more sensitive to rate increases than broadly syndicated loan borrowers.
The common thread is not that private credit is bad. It is that the asset class has scaled to systemic size without a full-cycle stress test, and that the data needed to see the transmission channels does not yet exist at fund or loan level.
Where it is contested.
The disagreement is real and should not be smoothed over. A second strand of analysis argues that private credit's structure is stabilising: long-locked capital, no depositor runs, losses absorbed by sophisticated institutional investors rather than the banking system. So a universal owner is choosing between two genuinely held positions: private credit as a fragile, untested concentration, or private credit as a shock absorber that moved credit risk off bank balance sheets to investors who can hold it. The honest reading is that the answer is empirical, depends on the next downturn, and cannot be known in advance — which is itself the planning assumption.
From the allocator's seat.
The commercial implication is not reduce the allocation. For most large owners the allocation funds a return target they cannot otherwise hit, and it is illiquid by design. The implication is that the work shifts from sizing the position to governing it. Concretely, an investment team should be able to answer, in writing, before the next committee meeting: what share of the book sits in semi-liquid, redemption-offering vehicles, and what the redemption mechanics are under stress; how much leverage sits between the fund and the underlying loan, including fund-level facilities; what the covenant quality actually is, deal by deal, not at the strategy level; and which banks are on the other side of the fund's credit lines. The same discipline travels to the AI-infrastructure book, where similar underwriting is being done under a different label.
What to watch next.
The FSB has proposed a core set of comparable metrics for authorities to track; watch for the first jurisdiction to adopt them, because that is when the data gap starts to close. Watch the next IMF GFSR for an updated read on valuations and interconnections. Watch the spread of semi-liquid vehicle launches, the FSB's clearest procyclicality concern. And watch the first credit event large enough to test a redemption gate — that is the moment the two contested positions stop being theoretical.
UAO Research. AI-assisted monitoring and drafting; reviewed and edited by the UAO editorial desk before publication. Not investment advice.