Sponsorship: Private Debt News reaches institutional investors, credit professionals, and LP decision-makers. Early sponsor rates available. Contact us here or reply directly to this email.
Mercer Advisors owed roughly $1.6 billion to KKR, Ares, BlackRock, and Apollo at S+450. Last week it priced a $1.65 billion TLB at S+275 and 99.75, taking 175 bps out of its margin and saving about $29 million a year.
Mercer is one credit, but the direction of traffic is what matters. JPMorgan and KBRA DLD count $19.5 billion of private credit refinanced into the BSL market this year against $9.2 billion going the other way. “If borrowers have the ability to access the broadly syndicated market today and it’s not a complicated financing, they are probably going to favor that market because it’s strictly a cost of capital conversation,” said DC Advisory’s Michael Moore. For a decade the pitch was that borrowers paid up for speed and certainty. Mercer’s CFO called the refi “a natural next step.”
The Fed hiked last week. CCCs pushed toward 14% for the first time in nearly three years. The Alternative Credit Council put 14.1% of direct lending borrowers below 1.0x coverage in the second quarter. Fitch’s default rate hit a record 6.3% in August.
Redemptions held their pattern. Morgan Stanley capped its $7 billion fund at 5% after 11.4% sought out, with two-thirds of requests from investors already blocked twice. Oaktree came in at 3.8%, under the cap for a second straight quarter. Jon Gray told NZZ that BCRED’s weakest 5% is marked around 60 against 95 for the senior book, and dismissed the critics. “The town criers are proclaiming a global crisis, and then nothing happens.”
Elsewhere, an H.I.G. BDC with 7% non-accruals put itself up for sale, KBRA found European mid-market borrowers running a 0.3% default monitor against 2.8% in the US with spreads 50 bps wider, and Fortress co-CEO Jack Neumark warned lenders off AI infrastructure FOMO. Blackstone is out raising $8 billion for it anyway.
Key Market Themes
1. Borrowers Head for the Exit
Mercer Advisors, the Oak Hill-owned wealth manager with $111 billion in client assets, priced a seven-year $1.65 billion TLB at S+275 and 99.75 last week, plus a $250 million DDTL for acquisitions. Goldman led. Proceeds retire about $1.6 billion of private credit debt from KKR, Ares, BlackRock, and Apollo’s MidCap that carried S+450. The savings come to roughly $29 million annually.
The flow is lopsided. Year to date, $19.5 billion of private credit has been taken out by BSL versus $9.2 billion moving the other way, per JPMorgan and KBRA DLD. DC Advisory’s Michael Moore framed it as a pure cost-of-capital decision for any borrower without a complicated story.
Speed was worth something once
Private credit’s premium was supposed to buy certainty and execution. When banks were retreating and the syndicated market was shut, borrowers paid the extra 150-200 bps without complaint. Mercer is a clean, growing, sponsor-backed business with no software exposure and no drama, and it just concluded the premium buys nothing. The $19.5 billion heading out is the good paper, because that’s what the BSL market will take. What stays in private credit is whatever can’t clear a syndication, and the average quality of the remaining book deteriorates with every Mercer that leaves.
2. The Hike Lands on 14% of Borrowers Already Underwater
The Fed raised last week. CCC yields climbed for a ninth straight session to 13.97%, with the premium over BB/B within three bps of its 2022 high. The Alternative Credit Council’s second-quarter update put 14.1% of direct lending borrowers below 1.0x coverage. Fitch’s US private credit default rate reached a record 6.3% in August, up from 6.1% in July. PIMCO’s Lotfi Karoui summarized the measurement problem in three words. “Estimates are all over the map.”
Carlyle’s Alex Chi pushed back, calling defaults “still within cycle averages” and arguing the volatility creates opportunity across opportunistic, ABF, and corporate lending. Carlyle’s non-traded BDC has seen requests at or below 5% for several quarters while taking inflows. Colleague Lauren Basmadjian noted AI-adjacent issuers are “generally price agnostic” and willing to give on terms.
The arithmetic on floating rate
A borrower at 0.9x coverage before the hike is at something worse after it. That’s 14% of the direct lending universe, and it grows with each 25 bps. The industry’s answer for two years has been PIK, which converts a coverage problem into a bigger principal balance. Chi is right that defaults sit within historical ranges, but the historical range was built during a period when rates fell after every shock. This is the first cycle where the Fed hikes into rising defaults. Karoui’s point is the honest one. When six metrics report six default rates, the comfortable number is the wrong one.
3. Gates Hold, Gray Shrugs, Oaktree Stays Under
Morgan Stanley’s North Haven Private Income Fund capped at 5% after 11.4% of shares sought out, essentially flat quarter over quarter. Nearly two-thirds of requests came from investors already limited in the prior two offers. Cumulative repurchases across three periods total about $479 million. Oaktree’s Strategic Credit Fund received 3.8% and paid in full, its second straight quarter under the cap, with repayments and July inflows exceeding the tender.
Jon Gray told NZZ that Blackstone expects inflows to return over time and put the weakest 5% of BCRED’s book at roughly 60 against 95 for the full senior portfolio. Operating income in the loan book grew double digits last quarter. Demand has shifted toward PE and infrastructure products for now. Franklin Templeton’s Jenny Johnson said the same thing at IPEM, that meeting redemptions above 5% trains investors to expect it.
Two-thirds of the queue is the same people
The Morgan Stanley detail is the one to sit with. Two-thirds of this quarter’s requests are investors on their third attempt. The backlog isn’t being refilled by fresh panic so much as recycled by the same holders who decided months ago they want out. That’s more stable than a run, but it also means the $15 billion queue clears only as fast as 5% a quarter allows. Gray’s 60-versus-95 framing is a reasonable description of a portfolio with a bad tail and a healthy core. It’s also an admission the tail exists and has already cost 40 points. Oaktree keeps demonstrating that investors who signed up for a distressed specialist behave differently from investors who bought yield.
4. A BDC Puts Itself Up for Sale
WhiteHorse Finance, the roughly $570 million listed BDC managed by H.I.G. Capital, formed a special committee to weigh strategic options including a sale. Almost 7% of the portfolio at cost sits on non-accrual. Shares traded more than 35% below NAV, and rising leverage forced the fund to pause its buyback in May. Soured loans include an Orangetheory franchisee and the maker of Hacky Sack. Shares rose 5.5% on the news.
“They are above the group average in the amount of credit losses and that explains why they are trading far below book,” said Oppenheimer’s Mitchel Penn. BlackRock’s TCPC is running a similar process after hiring advisers to sell its remaining loans.
The small BDCs go first
WhiteHorse follows TCPC as the second sub-scale BDC this year to conclude it can’t fix itself. Both are small, both carry non-accruals well above peers, and both trade at discounts that make raising equity impossible. That’s the trap for the middle tier. A BDC that can’t issue below NAV can’t grow, can’t refinance its way out, and can’t hold troubled loans through workout without leverage rising further. H.I.G. is a large manager with plenty else to do, and Penn’s point about capital allocation is the polite way of saying the parent has moved on. Expect more of these. The listed BDC market is where consolidation gets forced, because the price prints every day.
5. Europe Looks Cleaner, and Not by Accident
KBRA’s first head-to-head comparison of its US and EU/UK mid-market portfolios found the European book materially healthier on a point-in-time basis. The Middle Market Default Monitor sits at 2.8% for US borrowers versus 0.3% in EU/UK. Median coverage is 1.9x in Europe against 1.6x in the US. Median EBITDA margin is 19% versus 16%. Spreads on b- rated European loans run about 50 bps wide of US equivalents, despite 82% of European debt rated b- or better.
KBRA attributes the gap mostly to vintage. The median European loan was originated in 2024, the median US loan in 2023, and the US portfolio carries more older deals that have had time to sour. The study covers 2,226 US borrowers with $961 billion of debt and 465 European borrowers with $236 billion. KBRA expects the advantage to narrow, and warned that managers without local underwriting capability face weaker outcomes.
Wider spreads on better credits
The European premium is upside down by the logic of the last decade. Investors get paid 50 bps more to lend to companies with better coverage and lower default rates, because fewer lenders compete there and the legal patchwork commands a complexity premium. That’s a genuine mispricing for anyone with the infrastructure to underwrite across jurisdictions, which is why Arini raised $4 billion for a debut European fund and Apollo is building a €10 billion platform. KBRA’s vintage caveat is fair, and a 2024 European loan will look worse in 2027. But the US book is what a 2023 vintage looks like after a hiking cycle, and that’s the more useful data point for anyone still deploying into American mid-market paper at tighter spreads.
6. Fortress Warns Against AI FOMO as Blackstone Raises $8 Billion for It
Fortress co-CEO Jack Neumark told the Milken Canada summit that lending to data centers and GPUs carries the wrong asymmetry. “As a credit investor, you’re not getting paid for that upside and you’re stuck in the investment if it goes sideways.” His prescription is short duration, a hard view on residual value, and a path to exit or restructure before the technology moves. Trimontium’s Yaroslav Syzonov was blunter on price. “Is 8% really an attractive return?”
The money keeps flowing. Blackstone is seeking at least $8 billion for its fourth green infrastructure credit fund covering energy transition, data centers, and chip financing, after the prior vintage returned a 15% net IRR. Carlyle closed $2.3 billion for infrastructure credit, triple its predecessor. A CleanSpark data center fully leased to Meta sold $2.28 billion of junk at 8.25% on $10 billion of orders. SoftBank borrowed nearly $21 billion and wants another $10-20 billion this week. Apollo is in talks to upsize a SoftBank loan to $9 billion.
Fixed upside, open downside
Neumark’s argument is the one every lender in this space should be able to answer and most can’t. A 20-year Meta lease with a rent guarantee solves the counterparty question and does nothing for the residual value of hardware in year eight. Blackstone’s 15% IRR on the prior fund was earned on assets that appreciated through a boom. The next vintage is being raised into $1-1.5 trillion of annual capex that Alberto Gallo calculated requires hyperscalers to double revenues just to service. Basmadjian’s line that issuers are price agnostic and give on terms is the tell. Borrowers who need the money more than they care about the coupon are usually the ones you end up restructuring.
Deals of Note
Mercer Advisors - $1.65B seven-year TLB at S+275 and 99.75, plus $250M DDTL, refinancing ~$1.6B of private credit at S+450 from KKR, Ares, BlackRock, and Apollo
Real Brokerage - Fortress and Kennedy Lewis providing $550M at S+550 with a 3% floor to back the $880M Re/Max acquisition, taking out Morgan Stanley and Apollo bridge commitments
Blackstone green infrastructure credit - Seeking $8B+ for fourth fund; prior vintage returned 15% net IRR
Carlyle Infrastructure Credit Fund II - Closed $2.3B, triple its predecessor, ~$500M already committed
Partners Group - Exploring €800M continuation vehicle for loans from 2018 and 2020 credit funds
Crescent Capital - Sold $3.2B of stakes to Pantheon
Benefit Street Partners - $2.3B continuation fund led by Coller
Tahmoor Coal - $140M five-year senior secured loan from RRJ Capital to restart the Australian coking coal mine
CleanSpark - $2.28B junk bond at 8.25% for a Meta-leased data center, ~$10B in orders
WhiteHorse Finance - H.I.G.-managed BDC exploring a sale with ~7% non-accruals
The Reality Check
Mercer leaving 175 bps tighter is more consequential than any redemption headline this quarter. Redemptions are about who holds the paper. Refinancing is about whether the paper exists at all. When a clean sponsor-backed borrower with no complications decides the private credit premium buys nothing, the asset class loses its best credits to the bank market and keeps the ones the bank market won’t touch. That’s adverse selection running two to one in real time.
The hike turns a slow problem into a compounding one. Fourteen percent of borrowers were already under 1.0x before last week. Managers can PIK through another quarter or two, but PIK is a bigger principal balance on a company that couldn’t cover the smaller one. Karoui’s remark that default estimates are all over the map is the polite version of a harder truth, which is that the number everyone reports is the one that looks best.
Gray’s 60-versus-95 split is worth taking at face value. The core of BCRED is fine and the tail has already lost 40 points. That’s a manageable portfolio. It’s also a description of an asset class with a visible bad tail, and the question for every smaller fund is whether that tail is 5% of the book or, as at WhiteHorse, closer to 7% and growing. Sub-scale BDCs find out first because the screen tells them every day.
Europe paying 50 bps more for cleaner credit is a mispricing that won’t last, and Fortress warning about AI FOMO while Blackstone raises $8 billion for it is a disagreement residual values will settle around 2030. In the meantime the two-to-one refi flow, the record default rate, and a Fed hiking into weakness describe a market that hasn’t collapsed and isn’t healing. It’s slowly getting worse, one Mercer at a time.