What did the third-quarter private credit default index show?
A second quarterly decline at the headline, and a rise underneath it. Proskauer's Private Credit Default Index, which tracks senior secured and unitranche loans in the United States, printed 2.41 percent for July to September 2026, down from 2.51 percent in the second quarter (Proskauer, 27 October 2026).
This quarter's index covers 746 loans with $205.7 billion of original principal. By borrower size, the rate for companies with less than $25 million of EBITDA rose to 2.2 percent from 1.9 percent; for $25 million to $49.9 million it eased to 3.3 percent from 3.4 percent; and for $50 million or more it fell to 1.7 percent from 2.4 percent (same release).
Proskauer reads the decline as a sign of "continued resilience" and attaches a caution to it. "The headline number doesn't tell the whole story," said Stephen A. Boyko, partner and co-founder of the firm's Private Credit Group. "Trailing 12-month defaults rose, software and technology defaults climbed, and performance varied by sector and borrower size" (same release). In its first and second-quarter releases the firm had described software and technology default rates as stable (Proskauer, 27 April and 28 July 2026).
Where did the improvement in the default rate come from?
From the top of the market. Of the three size bands Proskauer reports, only the largest moved down by more than a tenth of a point: borrowers with $50 million or more of EBITDA went from 2.4 percent to 1.7 percent. The middle band was all but unchanged and the smallest rose (Proskauer, 27 October 2026).
The release does not say how much each band contributed to the headline, and we do not estimate it, because the index does not publish how many loans sit in each band. What the bands do show is direction: the average improved while borrowers below $25 million of EBITDA went the other way, their second rise in three quarters.
Two quarters of decline also need a date beside them. A year earlier the index printed 1.84 percent, and each band was lower than it is now: 1.6 percent below $25 million, 2.6 percent between $25 million and $49.9 million, and 1.2 percent at $50 million or more (Proskauer, 23 October 2025). On our arithmetic the headline is 0.57 points above its level of a year ago and every band is between 0.5 and 0.7 points higher, which is consistent with Boyko's remark that trailing twelve-month defaults rose.
Which borrowers have defaulted most often in the index?
Companies with $25 million to $49.9 million of EBITDA, in each of the last five quarters. That band printed 2.6, 3.6, 3.1, 3.4 and 3.3 percent from the third quarter of 2025 to the third quarter of 2026, above both the smaller and the larger band every time (Proskauer quarterly releases, 23 October 2025 to 27 October 2026). This quarter it is about twice the rate of the largest band, on our arithmetic.
Proskauer does not explain the pattern, and neither its releases nor this article can say why it holds. It runs against the common assumption that default risk falls steadily as a borrower grows. It is also not the only size reading in the market: a different dataset, with different bands and a different definition of default, finds stress concentrated among the smallest borrowers, as a companion piece sets out. The two are not comparable line for line.
The population behind the rates also changes. Over the year the index grew from 705 loans with $141.0 billion of original principal to 746 loans with $205.7 billion (Proskauer, 23 October 2025 and 27 October 2026); on our arithmetic the average original principal per loan rose from about $200 million to about $276 million. A rate can move because new loans enter the index as well as because old ones default, so one quarter's change in one band is a signal to watch rather than a verdict.
| Quarter | All loans | EBITDA below $25m | $25m to $49.9m | $50m and above |
|---|---|---|---|---|
| 3Q 2025 | 1.84% | 1.6% | 2.6% | 1.2% |
| 4Q 2025 | 2.46% | 1.7% | 3.6% | 2.4% |
| 1Q 2026 | 2.73% | 2.3% | 3.1% | 3.0% |
| 2Q 2026 | 2.51% | 1.9% | 3.4% | 2.4% |
| 3Q 2026 | 2.41% | 2.2% | 3.3% | 1.7% |
“A falling average and a rising rate in your own size band can both be true in the same quarter. Lenders read the band, and the trend in it. An owner should read the same line before the first conversation with a lender, not after it.”
What is a large lender's own book showing?
A rise, read from a lender's balance sheet rather than an index. At Ares Capital Corporation, a listed lender to private middle-market companies in the United States, loans on non-accrual were 2.4 percent of the portfolio at amortized cost and 1.4 percent at fair value at 30 September 2026, against 1.8 and 1.2 percent at the end of 2025 (Ares Capital, third quarter 2026 results, 27 October 2026).
Its chief executive, Kort Schnabel, described credit quality as stable, with "historically low levels of non-accruing loans and problem assets" (same release). Both things can be true: a non-accrual rate can be low by a lender's own history and still have risen by 0.6 points of cost in nine months, on our arithmetic.
New lending was slower in the same quarter. Ares Capital made $2.1 billion of gross commitments, against $3.9 billion a year earlier, to 25 new portfolio companies and 28 existing ones. The company attributed the more measured pace to a slower transaction environment, and reported a backlog of about $2.5 billion at 22 October and stronger activity entering the fourth quarter (same release).
What should an owner do when the average default rate falls?
Read your own band, not the headline. In our judgement a lender looking at a new loan, an add-on or an amendment for a company with $25 million to $50 million of EBITDA has that band's record in mind, and in this index it has been the highest for five quarters. Expect the questions that follow from it: covenant headroom, the cash forecast behind the headroom, and what happens to both if the next two quarters come in below plan.
Expect closer attention, and get ahead of it. Proskauer, whose Private Credit Group acts only for private credit investors, says close portfolio monitoring and early engagement with borrowers "matter more than ever" (Proskauer, 27 October 2026). For a borrower the same advice runs the other way: raise a problem before the quarter-end test raises it for you, with a forecast and a proposal attached. A request made early and with numbers is a negotiation; one made after a breach is a remedy.
Then watch two prints. Proskauer's fourth-quarter index (last year's came on 26 January) will show whether the smallest band's rise held, and Ares Capital's fourth-quarter results will show whether non-accruals kept climbing as originations picked up. A third quarter of headline decline with the smaller bands rising would say that the average and the borrowers below $50 million of EBITDA are moving apart.
As of October 2026
Sources: Proskauer Rose LLP, Proskauer's Private Credit Default Index Reveals Rate of 2.41% for Q3 2026, release, New York, 27 October 2026, read first-hand on proskauer.com, for the third-quarter 2026 rate of 2.41 percent against 2.51 percent and the second consecutive quarterly decline, the 746 loans and $205.7 billion of original principal, the rates of 2.2, 3.3 and 1.7 percent by EBITDA band against 1.9, 3.4 and 2.4 percent in the second quarter, the index's coverage of senior secured and unitranche loans in the United States, the availability of the full report only to the firm's direct lending clients, the Private Credit Group's exclusive representation of private credit investors, and the remarks of Stephen A. Boyko; the same firm's releases for the second quarter of 2026 (28 July 2026), the first quarter of 2026 (27 April 2026), the fourth quarter of 2025 (26 January 2026) and the third quarter of 2025 (23 October 2025), each read first-hand, for the earlier headline rates, band figures, loan counts and original principal in the table and for the description of software and technology default rates as stable in the first and second quarters of 2026. Ares Capital Corporation, Ares Capital Corporation Announces September 30, 2026 Financial Results and Declares Fourth Quarter 2026 Dividend of $0.48 Per Share, Form 8-K exhibit 99.1, 27 October 2026, read on SEC EDGAR, for loans on non-accrual of 2.4 percent at amortized cost and 1.4 percent at fair value at 30 September 2026 against 1.8 and 1.2 percent at 31 December 2025, gross commitments of $2.1 billion in the third quarter of 2026 against $3.9 billion a year earlier, new commitments to 25 new and 28 existing portfolio companies, the investment backlog of about $2.5 billion at 22 October 2026, and the remarks of Kort Schnabel. The year-on-year changes by band, the ratio of the middle band to the largest, the average original principal per loan and the change in non-accruals are our arithmetic. The companion article cited for a different size reading carries its own sources. The readings on lenders, amendments and timing are ours.
Corrections: factual errors are corrected on the page and the correction dated. Write to admin@kadenwoodcapital.com.
This position sits within our debt advisory practice.