Private credit is not breaking in the way markets normally expect. Large platforms still report mostly first-lien, floating-rate books, and non-accruals remain contained. Yet the first signs of strain are already visible in the structures around those assets: redemption caps, discounts to NAV, selective loan markdowns, and tighter bank financing. That matters because investors are starting to charge for liquidity mismatch, valuation lag, and old-cycle underwriting before headline defaults have done the work for them.

The Investment Question

Can private credit still be treated as a stable income diversifier when the first signs of pressure are showing up in redemption mechanics, loan marks, and bank financing rather than in broad reported defaults?

The cleanest way to misread the setup is to force private credit into a classic crisis template. If the asset class were already in a full bust, the evidence would look simpler: non-accruals would be surging, recoveries would be collapsing, and managers would be openly discussing impairments rather than liquidity management. That is not where the market sits today. The better description is a repricing phase. The underlying loan books have not yet broken broadly, but the wrapper around them is being stress-tested, and that alone is enough to change the allocation debate.

The Wrapper Gets Tested Before the Loan Book Does

The distinction starts with the semi-liquid funds. Ares Strategic Income Fund and Apollo Debt Solutions both hit withdrawal limits after redemption requests rose into the low double digits. BCRED also showed that even a flagship franchise can face strain before broad credit losses become visible: the fund posted its first monthly loss in more than three years and still needed an expanded repurchase limit plus sponsor capital to meet requests.

A semi-liquid fund is only as liquid as its tender terms, financing lines, and remaining investors. Once withdrawal requests run through the cap, investors discover that the wrapper behaves much more like locked capital than a high-yield cash alternative. None of that proves the underlying assets are broadly impaired. It does prove that periodic liquidity only feels comfortable while investors are willing to accept slowly moving marks.

Legacy Books Can Feel Fine and Still Be Worth Less

The deeper issue sits behind the wrapper: legacy-book repricing. A private loan does not need to default to be worth less. If the market now requires a higher return for the same risk than it did when the loan was written, fair value falls even if the borrower keeps paying on time. Floating-rate coupons help against base-rate duration risk. They do not protect against spread widening, a higher illiquidity premium, weaker confidence in recoveries, or a more skeptical view of the sector itself.

Software has become the clearest pocket where this repricing is visible. The Medallia markdown mattered not because one name defines the asset class, but because it showed how a paying loan can still be revalued sharply when assumptions about cash-flow durability and AI disruption change. Once that happens, the effect does not stay inside one position. It feeds into bank financing, public-market sentiment, and the credibility of other marks across the complex.

Public markets are already leaning into that question. BXSL, OBDC, and ARCC all trade below NAV even though the larger listed BDCs still show manageable non-accruals and mostly senior-secured portfolios. Those discounts do not prove the internal marks are wrong. But when listed vehicles trade at meaningful discounts while semi-liquid funds are gating withdrawals and selected loans are being written down, the signal is difficult to dismiss.

Banks, Sponsor Finance, and the IRR Problem

Private credit is not a closed loop of funds lending to companies. Banks provide financing, warehouse assets, write NAV facilities, and increasingly help clients trade around the stress. Once banks financing the ecosystem begin re-marking collateral, private credit stops looking like a one-way yield product and starts looking like an asset class the broader market is actively repricing.

That has immediate implications for private equity because private credit is now central to sponsor finance. If lenders demand wider spreads, lower leverage, and tighter structures, sponsor math changes immediately. A higher debt burden compresses equity IRRs unless sponsors can offset it through lower entry multiples, stronger EBITDA growth, or optimistic exit assumptions. The result is not the end of private-equity transactions. It is a more selective deal market with less room for error.

The Allocation Call

2026 looks more like a bifurcation year than a collapse year. The weak point is broad exposure to legacy semi-liquid vehicles that were sold as smooth income while carrying older books written in a more forgiving spread regime. The more interesting opportunity may be fresh-vintage direct lending written at wider spreads, tighter terms, and more realistic assumptions.

Our conclusion is straightforward: lower private-credit allocation here by cutting exposure to semi-liquid wrappers and older books where valuation lag can do as much damage as realized defaults. That is not a call to abandon the sector. It is a call to separate yesterday's exposure from tomorrow's opportunity. If spreads widen further and structures reset, new-vintage direct lending may become one of the better parts of the credit market. For now, the market is still discovering which products were built on durable underwriting and which relied too heavily on the illusion of stability.