This paper has spent the week on what a private credit valuation is worth. A mark is an opinion until somebody has to sell. A PIK amendment moves a struggling borrower out of the statistic that counts non-payers. A secondary sale is the only place a mark meets a bid, and it only prices the stakes somebody chose to offer.
All of those resolve, eventually, into one number: how many cents come back when a borrower defaults.
Moody's currently puts expected ultimate recoveries on first-lien senior secured loans at about 68 percent. The long-run historical average is in the high seventies. Second lien runs 30 to 45 cents, with sector-dependent outcomes anywhere from 10 to 60.
The covenant is the mechanism
Recoveries are falling and the reason is not that lenders became worse at pricing risk. It is that they lost the instrument that told them when to act.
A financial maintenance covenant is tested every quarter whether or not anything has happened. Leverage above a threshold, or coverage below one, and the borrower is in breach — which does not mean the loan is called, but does mean the lender is at the table with leverage while the business still has value. That is the entire function: it is a tripwire, and it fires early.
Between 90 and 95 percent of broadly syndicated loans currently being issued have no such covenant. An incurrence covenant tests only when the borrower does something — raises debt, pays a dividend — so a company that simply deteriorates quietly, doing nothing, trips nothing.
The lender's first enforceable moment therefore moves from the quarter the numbers went wrong to the quarter the borrower missed a payment. Everything between those two points is time in which enterprise value erodes with no creditor able to intervene.
That is why the loss shows up in severity rather than frequency. Default rates can look unremarkable while recoveries decline, and the two statistics are usually reported by different people.
And the delay has been extended again
Read the PIK numbers against this and the picture sharpens.
A borrower amended from cash interest to payment-in-kind is not in default and does not appear in non-accruals. The clock that would have started running has been reset by agreement, and the principal grows while it is reset. Every quarter bought that way is a quarter added to the interval between deterioration and intervention — the same interval covenant-lite already lengthened.
The two mechanisms compound. One removed the early tripwire; the other postpones the late one.
What the pipeline looks like
KBRA's fourth-quarter middle-market compendium ran 3,649 borrower assessments. Median interest coverage was 1.5 times. Twenty-five percent of borrowers were below 1.0 times — which is to say a quarter of them were not earning enough to cover their interest. Nineteen percent reported declining sales, 22 percent declining EBITDA, and multilevel downgrades rose 2.9 times quarter on quarter.
A borrower below 1.0 times coverage is funding interest from somewhere other than operations: a revolver, a sponsor equity cheque, or an amendment. None of those is a going concern and all of them are invisible in a default rate.
Where the mezzanine went
One figure sits oddly beside the rest and explains a structural change. Mezzanine fundraising fell 82 percent, to $6.6bn in 2024, while mezzanine pricing runs 10 to 14 percent all-in with effective returns of 16 to 22 percent once warrants are counted.
Capital does not usually flee a 20 percent return. What happened is that unitranche lending absorbed the space between senior and equity — one instrument, one lender, no intercreditor negotiation. That is faster to execute and it removes the junior creditor whose job was to be wiped out first and who, in the process, used to do a great deal of the arguing about enterprise value on the way down.
The 35 to 50 percentage point gap between first and second lien recoveries is what that layer was absorbing. Where it no longer exists, the outcome is more binary.
The number that would tell you
Not the default rate. Watch first-lien recoveries on middle-market loans that defaulted after 2022, as they resolve, against the pre-2015 cohort that still had maintenance covenants.
If the gap holds at ten points or widens, covenant-lite has permanently repriced what senior secured means, and a decade of loans were underwritten against a recovery assumption that no longer applies. The loans priced for the old number. The workouts will settle at the new one, and the maturity wall is where those two facts meet.
The first-lien recovery estimate of about 68 percent against a long-run average in the high seventies, second-lien recoveries of 30 to 45 cents with a 10 to 60 percent range, and the 35 to 50 percentage point gap between lien positions are attributed to Moody's Ultimate Recovery Database and S&P Global Ratings recovery studies covering 1987 to 2026, as compiled by ABF Journal in September 2026. The estimate that 90 to 95 percent of broadly syndicated loans in current issuance lack financial maintenance covenants, the mezzanine all-in pricing of 10 to 14 percent with effective IRRs of 16 to 22 percent, and the 82 percent fall in mezzanine fundraising to $6.6bn in 2024 attributed to PitchBook are from the same compilation. The KBRA Q4 2025 Middle Market Borrower Surveillance Compendium figures — 3,649 assessments, median interest coverage of 1.5x, 25 percent of borrowers below 1.0x, 19 percent declining sales, 22 percent declining EBITDA and multilevel downgrades up 2.9 times quarter on quarter — are as cited there. The analysis is our own.
Topics marketsprivate credit
