Private credit is under scrutiny. Is it warranted?

In the past sixty days, Apollo capped redemptions on its $25 billion flagship credit fund. Morgan Stanley locked the doors on a $26 billion vehicle. Blue Owl quietly restructured how investors can exit—switching from quarterly buybacks to drip-feed capital returns. Fortune ran a headline calling it a “$265 billion meltdown.”

If you own private credit—or you’re considering it—you’re probably wondering whether this is the beginning of something ugly.

Here’s the honest answer: Some of the panic is warranted. Most of it isn’t. And the distinction matters enormously for what you should do next.

The Numbers Behind the Headlines

Let’s start with what’s actually happening. The Proskauer Private Credit Default Index hit 2.46% in the fourth quarter of 2025, up from 1.84% the prior quarter. That’s a meaningful jump. But the scarier figure comes from CNBC’s analysis: Once you include distressed exchanges and liability management exercises—the financial equivalent of rearranging deck chairs—the “true” default rate is closer to 5%. Default events rose 78% year-over-year in 2025.

Those numbers deserve your attention. They do not, by themselves, deserve your panic.

Private credit is a $3.5 trillion market that expanded at an extraordinary pace after the 2008 financial crisis, when banks pulled back from middle-market lending and institutional investors rushed to fill the gap. When that much money chases that many deals, underwriting standards slip. That’s not a theory—it’s physics. The stress we’re seeing now is the market doing exactly what markets do: Correcting the excesses of the previous cycle.

The AI Wild Card

But there’s a newer wrinkle that makes this cycle different from any previous credit stress test: artificial intelligence.

Here’s the connection most people miss. A huge chunk of private credit—roughly 26% of direct lending portfolios—is loans to software companies. Many of these are SaaS businesses acquired by private equity sponsors, leveraged up, and financed with floating-rate debt that was underwritten when rates were at 5–6%. Those same loans now carry rates of 10–12%. Outstanding loans to SaaS firms alone grew from $8 billion in 2015 to over $500 billion by end-2025.

That alone would be painful. But AI introduces a second problem. If AI tools can replace three software subscriptions with one, or if they let companies build in-house what they used to buy off the shelf, then the revenue growth assumptions baked into those loan covenants may never materialize. Morgan Stanley has warned that default rates in software-heavy portfolios could reach 8%. UBS puts the “severe scenario” at 15%.

This doesn’t mean every software-backed loan is going bad. It means lenders who understood which software companies had durable competitive advantages—and which were just selling seats—are going to look a lot smarter than those who underwrote to a revenue multiple and called it a day.

The Liquidity Illusion

The gating headlines are dramatic, but they’re also somewhat misleading. Here’s why: The underlying loans in private credit were never liquid. They’re negotiated, bespoke instruments held for years. The problem isn’t that private credit became illiquid—it always was. The problem is that Wall Street sold it in wrappers that implied otherwise—interval funds, evergreen vehicles, non-traded BDCs.

When investors tried to redeem all at once in Q4 2025, redemption requests as a share of fund NAV nearly tripled. Fund managers had two choices: sell assets at fire-sale prices, or gate. They chose the gates. This isn’t a sign that private credit is broken. It’s a sign that the packaging was always a little dishonest about what “semi-liquid” actually means when everyone heads for the exit at the same time.

The Valuation Question

There’s one more issue worth naming directly. I don’t think it helps anyone to be coy about this: private credit valuations lag reality.

These assets are marked quarterly using models, not markets. When public credit spreads widen — as they have—the signal is clear that risk is being repriced. But private credit marks often take months to catch up. The DOJ has publicly flagged “creative” valuation practices. The SEC is probing rating agency independence. Analysts expect significant write-downs in Q2 2026 earnings.

This doesn’t mean the assets are worthless. It means that reported stability and actual stability are two different things right now, and investors should understand the difference.

Due Diligence in Direct Lending

Here’s where I’ll get specific, because I think investors deserve more than “manager selection matters”—even though it does.

When we built our private credit allocation, we weren’t chasing the highest yield. We were underwriting the underwriters. We prioritized managers with deep credit backgrounds—teams that had restructured loans in 2008, not teams that had only ever deployed capital in a zero-rate world. We focused on strategies with genuine structural protections: overcollateralization, cash flow waterfalls, and senior positioning in the capital stack. And we deliberately avoided the crowded trade in sponsor-backed software lending that’s now making headlines.

None of that makes our allocation immune to this cycle. Private credit is a credit product, and credit products have bad years. But there’s a meaningful difference between managers who prepared for a storm and managers who assumed permanent sunshine.

The Bottom Line

Private credit is not collapsing. It is correcting—and for the first time, doing so publicly enough that retail investors can see the process. The defaults, the gating, the valuation questions—these are features of a maturing asset class, not bugs in a failing one.

But this is also the moment where the gap between good managers and average ones becomes a canyon. The next twelve months will be messy. Dispersion will widen. Some funds will gate again. Some borrowers will restructure. And when the dust settles, the investors who chose discipline over yield will be the ones still standing.