Private Credit Default Rates: What the Data Actually Shows

There is no single private-credit default rate. The published figure and the real one disagree by a factor of three, because amendments, PIK and quarterly marks sit between distress and an actual default.

Key takeaways

• There is no single private credit default rate. For the same period, Fitch reads 6.0% (US, April 2026), Moody’s 1.6% to 4.7% (US, 2025), KBRA 2.0% by value (US middle market, Q4 2025), and Cliffwater’s realised US losses about 1.02% a year. Europe runs milder still on the same lens: KBRA forecast a 1.25% European default rate for 2025, up from 0.4% in 2024. The gap is definitional, not an error: Fitch includes deferrals and PIK conversions, while KBRA and Cliffwater count narrower, realised events.
• Recoveries are the real edge, and they are regime-dependent. First-lien direct loans have historically recovered more than senior unsecured high-yield bonds, so a higher default rate can still mean a lower loss. But how much you get back depends on the insolvency regime the loan sits under, which differs by jurisdiction, and KBRA data already shows US direct-lending recoveries (~53%) running well above European ones (~40%).
• The reported rate understates stress by construction, wherever the fund lends. Amendments, extensions and PIK keep struggling loans performing on paper, and private loans are marked quarterly by a model, not daily by a market.
• Watch PIK, not the headline. Rising PIK income and non-accruals lead reported defaults; the default rate is the lagging signal.

There is a comforting story about private credit, and a truer one. The comforting story is that direct lending barely defaults: the long-run indices show credit losses of about a percent a year, well below high-yield bonds, so the $1.7 trillion that has poured into the asset class since 2020 looks safe. The truer story is that “the private credit default rate” is not one number. It is at least four, they disagree by a factor of three, and the gap between them is the most important thing to understand about this market.

The actual default and recovery figures from the main providers sit side by side below, along with why they disagree. The reader’s usual assumption, that a low reported default rate means low stress, is where this goes wrong. Reported defaults can be low at the same moment borrower stress is rising, because the mechanisms that keep a loan out of the default column, amendments, extensions and payment-in-kind interest, are the ones a private lender controls. The data points one way: the headline number is real but partial, and the more useful signal sits in the spread between providers and in what is happening to PIK.

The figure that started the scare

The number doing the rounds is Fitch. Its US Private Credit Default Rate hit 6.0% on a trailing-twelve-month basis in April 2026, a record since the index began in August 2024, after touching 5.8% in January 2026. Its sub-index for privately monitored borrowers, the smaller and weaker companies, ran hotter still, at 9.4% in January 2026.

Six percent, rising, and a record reads like a market cracking. It is not, or not yet, and the reason is buried in how Fitch defines a default. A large share of the events it counts are not missed cash payments. They are interest deferrals and conversions to payment-in-kind, where the borrower stops paying cash and the lender rolls the interest into the loan balance instead. Fitch counts that as distress, correctly, because a company that cannot pay its coupon in cash is in trouble. But it is a different animal from a bankruptcy, and it is why Fitch’s number sits so far above everyone else’s.

Why the providers disagree

Now put the other measurements next to it. Moody’s, looking at full-year 2025, gives a range rather than a point: private credit defaulted at between 1.6% and 4.7% depending on whether you count distressed exchanges. A threefold spread inside a single provider’s own work is the cleanest proof that the “default rate” is a choose-your-definition number, and Moody’s warns that improving averages mask dispersion, with distressed restructurings making up roughly 65% of all defaults in 2025.

KBRA, which tracks direct-lending deals loan by loan, lands lower: a US middle-market rate of 2.0% by loan value and 3.4% by borrower count in Q4 2025. And Cliffwater, whose index reconstructs realised experience back to 2004, reports the mildest picture of all: average annual realised losses of about 1.02% a year across 2005 to 2023, a touch above leveraged loans and below high-yield bonds.

The numbers so far are all American, because that is where the largest indices sit, but this is a global asset class and Europe reads on the same lens. KBRA’s European arm, tracking the region’s direct-lending deals loan by loan, forecast a 1.25% European default rate for 2025 by volume, up from 0.4% in 2024. Low in absolute terms, but rising fast off a very quiet base, and the same direction of travel as the US measures. The lens travels; the level and the workout do not, which is the point of the next section.

So the same asset class, in the same window, runs anywhere from roughly 1% in Europe to 9.4% at the weak end of the US market, depending on who is counting and what they count. That is not a data error; it is the point.

Default and recovery across the providers (assembled)

The table below pulls the main measurements into one place, with the as-of date and the definition each provider uses. No cell is estimated; every figure is sourced above and at the foot. Read down the “what it counts” column and the disagreement stops looking like noise.

Two readings jump out. First, recovery is what makes private credit look better than its default rate suggests. A direct lender is often the only lender, or one of a small club, sitting at the top of the capital stack with tight covenants and a direct line to the borrower. When a deal goes wrong it can restructure privately rather than fight through a formal insolvency, and US first-lien direct loans have historically recovered around 54 cents on the dollar, against roughly 40 for senior unsecured high-yield bonds. A higher default rate at a higher recovery can still mean a lower loss. That is why Cliffwater’s realised-loss figure is so low even as Fitch’s climbs.

But recovery is not a universal constant, and this is where a global reader has to be careful. What a lender gets back when a borrower fails depends on the insolvency regime the loan sits under, and that regime changes with jurisdiction. A workout under US Chapter 11 (debtor-in-possession, court-supervised reorganisation), a UK administration or restructuring plan, and the various continental European regimes are different machines with different creditor rankings, timelines and outcomes. The gap is visible in the data already: KBRA puts US direct-lending recoveries near 53% against roughly 40% in Europe on its 2024 defaults, a small sample but a real divergence. So a recovery number is only meaningful once you know where the fund lends and under whose insolvency law it would restructure.

Second, the caution the headline skips: those recovery figures rest on thin, benign history, and there are early signs they are slipping. US middle-market first-lien recoveries have fallen towards around 66% on a trailing five-year basis, converging with broadly syndicated loans, per S&P Global data, below the figure the asset class long assumed. And the market has never been through a severe downturn at its current size, in any jurisdiction. The IMF’s warning is explicit: in an adverse scenario it expects a delayed realisation of losses followed by a spike in defaults and valuation markdowns. A 54% recovery is an average of good years, under one regime.

The mechanism that understates stress

Here is where the low headline numbers earn their asterisk. A public bond trades every day, so distress shows up in the price. A private loan is valued by the manager, quarterly, using a model rather than a market. The IMF puts it bluntly: private credit loans are “marked to model”, suffer from stale valuations, and managers “may be incentivized to delay the realization of losses” while they raise new funds. The US Federal Reserve made the milder version of the same point, noting reported defaults have been low versus the syndicated loan market in a “relatively opaque sector” where scarce data makes risk hard to assess.

When a private borrower struggles, the lender has options a bondholder does not, and those options work the same way in London, New York or Frankfurt. It can amend the covenants, extend the maturity, or let the borrower pay in kind. Each keeps the loan performing on paper; none means the borrower is healthy. This is what analysts mean by a “shadow default”, and it is a feature of how private lending works, not of any one country’s rules. Lincoln International, which values thousands of private loans, found PIK in about 11% of 2025 deals, most of it “bad PIK” that borrowers were forced into, giving what it calls a shadow default rate of roughly 6.4%, more than double the 2.5% of late 2021. That is stress the headline rate does not capture.

So PIK is the number worth watching, not the reported rate. PIK income across public business development companies, the listed wrappers around direct-lending portfolios, kept rising into 2026 on Moody’s data. When Moody’s downgraded one large BDC in early 2026, it flagged that PIK income had reached 14.7% of the fund’s investment income against a 6.3% peer median, with non-accruals at 5.5%. PIK leads reported defaults, because a borrower switches to PIK before it formally misses a payment. Watch only the default rate and you are watching the lagging indicator.

None of this makes private credit a bad asset. It makes the reported default rate an incomplete one. The number is not wrong; it is early, and it is smoothed.

What the data actually shows

Assemble it all and the honest conclusion is narrower and more useful than either the “it barely defaults” story or the “6% and rising” scare. Realised losses have genuinely been low, around a percent a year over two decades, on high recoveries and a lender’s control over the workout. But the reported default rate understates current stress by construction, because amendments and PIK sit between distress and default and because private marks move slowly. The spread between Fitch at 6.0% and KBRA at 2.0% is not a contradiction to resolve; it is a range to hold in your head, and the direction of travel across every provider is the same as higher-for-longer rates grind on floating-rate borrowers.

For a serious investor, anywhere in the world, that reframes the question. The number to interrogate is not the fund’s reported default rate, the flattering one, but its PIK share, its non-accruals, and how its manager has valued troubled positions, plus the insolvency regime its loans would restructure under if they failed. Those questions apply wherever the fund lends. The default rate is where the stress finally shows up, not where it starts.

FAQs

What is the private credit default rate right now?

It depends on the provider, their definition and the region. For the US, as of early 2026, Fitch put it at 6.0% (its widest measure, including PIK conversions and deferrals), Moody’s at 1.6% to 4.7% for 2025 depending on whether distressed exchanges are counted, KBRA at 2.0% by loan value, and Cliffwater’s realised losses at about 1.02% a year over 2005 to 2023. Europe reads milder on the same lens: KBRA forecast a 1.25% European default rate for 2025, up from 0.4% in 2024. All are current and all are correct; they measure different things in different places.

Why is Fitch’s number so much higher than the others?

Because Fitch counts more as a default. Its index includes interest deferrals and conversions to payment-in-kind, not just missed payments and bankruptcies. Those events signal real stress but are not a borrower failing outright, which is why narrower measures like KBRA’s and Cliffwater’s read lower.

Are private credit defaults understated?

The reported rate can lag actual stress. Lenders can amend covenants, extend maturities or let a borrower pay in kind, all of which keep a loan performing on paper, and private loans are valued quarterly by a model rather than priced daily by a market. Neither is wrongdoing, but both mean the headline rate moves late, which is why analysts watch PIK and non-accruals as earlier signals.

Does a higher default rate mean higher losses, and is the asset class safe?

Not automatically, because losses depend on recovery too, and first-lien direct loans have recovered more than high-yield bonds, so a higher default rate at a higher recovery can still produce a lower loss. But recovery is not universal: it depends on the insolvency regime the loan restructures under, which differs by jurisdiction (US Chapter 11, UK administration or a restructuring plan, and the various European regimes are different machines), and KBRA data already shows US recoveries running above European ones. And that history is mild, short and benign everywhere: the market has not been tested by a severe downturn at its current size, S&P data shows recoveries slipping, and the IMF has warned losses could be realised slowly then spike alongside valuation markdowns because private credit is not marked to market daily. This is general information and analysis, not personal financial advice.

Next read

• The pillar guide: Private Credit.
• The mechanism behind the numbers: PIK Toggle Explained: The Hidden Risk In Private Credit.
• How it compares to listed debt: Private Credit vs Public Credit.