A 25-basis-point gap now separates private credit's two senior funding markets

Private credit's senior risk now has two prices, and this week supplied them: Oak Hill's seasoned CLO reset cleared its AAA debt under SOFR plus 120, while Antares's upsized $850 million private credit CLO—its 2026-4 deal—priced AAA at 145, leaving at least twenty-five basis points between them. One vehicle had already been sold to investors; the other brought new collateral into a market whose appetite for private credit remains the open question.

A reset reprices liabilities on a vehicle whose assets the market has already seen, so a senior buyer under 120 is underwriting a pool that has traded once rather than a stack assembled this quarter, and the manager funds the same book more cheaply without selling a loan to do it. A sub-120 senior note print makes repricing a seasoned CLO cheap enough to widen the set of managers who can fund existing books through resets, and admission to that channel is priced off one thing: how long a manager has been lending.

Antares supplies the other end: new-issue private credit collateral still has liability buyers, and the upsized 2026-4 shows what they charge for a pool no one has held before. Both prints work, but only one can be repeated by the same manager against the same collateral.

Strip the labels and the two prints answer different questions. A senior holder in a seasoned vehicle buys a coupon with no equity upside and no origination story attached; what can go wrong is credit, and that credit has already been through a public underwriting once. The same buyer in front of new private credit collateral has to take a manager's word for how the loans were marked before anyone else owned them, and a spread in the twenties is a narrow price for that distinction—narrow enough to show the market is discriminating within private credit rather than turning away from it.

The second funding market does not reprice through a vehicle at all, and Ares Capital's unsecured note this week—printed weeks after its second quarter closed—reads as a funding test whose coupon will show what balance-sheet trust costs. An unsecured lender to a BDC sits behind everything: the loan book, the marks, the franchise that generated them, with no ring-fenced pool to dissect and no collateral test to run. Where the CLO buyer at under 120 owns a defined set of loans, the unsecured buyer owns the firm.

The Ares print carries more weight than its size because of who is doing it: a cost-of-capital call from a sector giant gets read as a temperature on the BDC model, and coming to market weeks after the quarter closed suggests the decision was taken with a freshly marked portfolio in hand rather than in anticipation of one. If an issuer of that scale pays up for unsecured money, smaller BDCs pay more; if it clears tight, the balance-sheet route is cheaper than the non-accrual trend alone would suggest.

That trend is why the route needs defending: the median non-perpetual-life BDC now carries 2.75% of its investments at cost on non-accrual, up from 1.81%. The deterioration is selective enough that manager selection is the only durable answer—easy for anyone who can see the portfolio, hard to act on from outside a credit agreement.

There is also a cost that will not show up in the headline coupon. A BDC earns the difference between what its assets yield and what its liabilities cost, and an unsecured claim on a whole balance sheet ought to price wider than a secured claim on a defined pool—senior CLO buyers took under 120 on exactly that kind of pool. Whatever the Ares coupon turns out to be, it comes out of net investment income before it reaches a shareholder, because the balance-sheet route claims the whole franchise rather than just the loans.

One distinction deserves care before anyone carries the non-accrual number across the asset class: the 2.75% median describes BDC portfolios, while this week's AAA spreads describe CLO collateral drawn from broadly syndicated loans and private credit assets assembled for the vehicle. Whether the stress now showing up in BDC marks is present in the pools being repriced is a question this week's coverage does not settle, and the two datasets are measuring different books.

Put the week together and private credit financing has split in two: public CLO liabilities, where a seasoned book reprices below SOFR plus 120 and a new one pays 145, and debt raised on a BDC's own balance sheet, where the lender is paid for a franchise and a set of marks. The two routes price different things, and a manager's funding cost now turns on whether it owns a book old enough to reset—a fact about the firm rather than the loans it holds this quarter.

Asset formation keeps pushing against that constraint. ICG's Europe Fund IX closed at its €12 billion hard cap, while Nest has put £650 million to work across private credit in five months without letting the asset class exceed four per cent of net asset value; ICG has also hired two structuring specialists from Schroders and Antin to build infrastructure debt tranches for a $126 billion platform. Each new book is a liability that has to be funded somewhere, and the manager who funds at 145 instead of below 120 pays that difference on every draw.

The reset route became a structure rather than an opportunity because so much of it printed at once: a seasoned reissue, a new-issue private credit CLO, Sona's second all-credit print, Eagle Point's move into infrastructure, and a euro hybrid in the October pipeline amount to several simultaneous bids for private credit collateral, each testing different terms. Managers without a seasoned book are not shut out of that market—they pay a different price, and in the hybrid's case they wait for a structure that has not yet been priced.

Blackstone is preparing a euro-denominated hybrid CLO for an October print that blends private credit with broadly syndicated loans, setting up a test of what it costs to put private assets under public CLO pricing. Sona's second all-credit print and Eagle Point's move into infrastructure collateral widen the same question: how far the CLO structure stretches past broadly syndicated loans before investors stop treating it as a CLO at all.

The answer decides how long this week's split lasts: if Blackstone clears its euro print anywhere near Oak Hill's level, then senior buyers are charging for structure and disclosure, and a hybrid can supply both without waiting a decade for seasoning. A market that pays up only for a track record taxes new managers permanently, and the hybrid is the structure asking whether that is what is happening.

None of the resets work without a liquid loan tape underneath them. PWD's tracking puts trailing loan trading volume at $978 billion, and the second-quarter dip that followed a record is different from a market losing its bid; the primary market says the same, with flat first-half volume on larger deals and Audax and Churchill tied at No. 1 among agents. Fewer, bigger loans are easier to price and trade, and tradable collateral is the precondition for repricing a vehicle that already owns it.

Watch the coupons. Ares's unsecured print is the first public price on the balance-sheet route, arriving against a median BDC non-accrual rate that has risen from 1.81% to 2.75%; the coupon will show whether the selectivity that underpins a franchise-level unsecured claim is real or a story told from inside the portfolio. Blackstone's euro print is the second price. If the two land close together, the gap was about the calendar; if they stay apart, the split is about the age of the collateral, and that is the bill the October print will present on every deal after it.