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.
In March, with Blue Owl’s stock collapsing and redemption requests piling up, Doug Ostrover gathered the firm and told staff the trouble came from “vicious” media coverage and rivals “always kind of on the prowl.” A recording of that meeting surfaced this week. The stock is down about 67% from its January 2025 high and sits near all-time lows. Investment bankers have been meeting the founders to pitch a full or partial sale of the business.
The numbers underneath the meeting are worse than the mood. Clients asked to pull 17% from Blue Owl Credit Income Corp last quarter, more than most competitors saw. OBDC II, the fund Blue Owl wound down by selling $600mm of loans in February, including a block to its own insurer Kuvare, reported its worst loss since the height of Covid. Having returned ~43% of capital by August through sales and repayments, what remains has concentrated the problems. Roughly 10% of the remaining loans are in default and another 20% are underperforming.
The strategic pivot isn’t producing the numbers either. Institutional fundraising fell 22% in the second quarter and only rose as a share of the total because individual flows dropped about 60%. Senior staff take roughly half their pay in deferred stock. CFO Alan Kirshenbaum’s slide for the all-hands showed KKR’s 38% drop in 2015 and its subsequent recovery. “We unfortunately know how that feels,” he said.
Blue Owl’s retail franchise was built on exactly the wirehouse concentration the industry diagnosed at Greenwich last week. Ostrover called individual investors “our hidden gem” in 2020 and raised about $50bn from them over three years. This week’s WSJ piece is what the gem looks like from the other side.
The rest of the week pointed the same direction. JPMorgan and Goldman had to push Volta’s $5bn AI loan to S+650 at 95 to clear it, a yield past 11%, while CoreWeave’s bonds trade above 15%. TCW cut its credit exposure to a 30-year low as CCCs sit at distressed levels. Goldman and UBS are reviewing fund-finance lines to Australian private credit managers. And a new academic paper found that a 10% wider discount to NAV at a listed BDC predicts a materially higher chance of a NAV markdown within three quarters, which is a problem for a sector trading around 82%.
Key Market Themes
1. Inside Blue Owl’s All-Hands
The March meeting came after Ostrover was flooded with employee questions. He acknowledged mistakes, mostly in public relations, and framed the pressure as a byproduct of relevance. Kirshenbaum’s presentation offered no crisis plan, just the KKR 2015 comparison. Some current and former employees told the WSJ the firm leaned too hard on wealthy individuals for growth and has struggled to respond. Costs from hiring, junkets, and pro-sports endorsements to build the retail machine drew resentment from other divisions when expense cuts arrived in mid-2025.
Departures include a Tokyo managing director to Benefit Street, an institutional MD to PIMCO, and a private-wealth principal to KKR. Blue Owl says attrition of 5.6% annually is below industry average. The firm’s larger, newer funds carry below-average default rates, and the company says the OBDC II sales delivered “a result for the investors that is far beyond what any other equivalent product has accomplished.”
The gem and the kryptonite
Blue Owl’s $50bn retail raise from 2022 to 2025 was the envy of the industry and it was also the single-channel concentration risk that every executive at Greenwich just finished explaining. The firm didn’t do anything its peers weren’t trying to do. It just did more of it, faster, through fewer pipes. Bankers pitching a sale tells you how the Street values the franchise now. A 67% drawdown with half of senior comp in stock means the firm’s own people have lost the most, which is why the all-hands leaked. The OBDC II residual at 10% default and 20% underperforming is the piece that should bother anyone still holding OCIC shares. It’s what a wound-down Blue Owl fund looks like once the good loans have been sold to pay out the first wave of exits.
2. Volta Clears at 11%, Which Is the Point
JPMorgan and Goldman sold Volta Infrastructure’s $5bn loan at S+650 and 95, up from initial talk of S+625 to 650 at 97 to 98. The all-in yield topped 11%. The banks added call protection and a ratings-based step-down on the coupon. Volta, a neocloud founded this year by two ex-Brookfield infrastructure executives, is financing a Norway data center run with bitcoin miner Bitdeer, with Anthropic already signed to a six-year compute deal. The package splits into a $1.3bn term loan to fund lease payments and a $3.7bn chip loan for about 36,000 Nvidia GPUs.
CLO managers balked at an unrated, debut, $5bn deal in a volatile window. CoreWeave’s bonds crossed 15% this week with its credit risk gauge at an all-time high. The banks had already curbed their own exposure.
What the chip loan costs now
A year ago, chip-backed paper with a hyperscaler or frontier-lab tenant priced inside 6%. Volta’s $3.7bn chip loan just cleared at an 11%-plus yield with a six-year Anthropic contract attached. That’s not a reflection of Anthropic’s credit. It’s the market putting a number on 36,000 GPUs sitting in Norway in 2030, and that number is low enough that lenders want a 500 bps premium over what the same trade fetched in 2025. Carlyle’s warning about AI concentration two weeks ago was theoretical. Volta is the price discovery.
3. Daily Marks Get Called an Illusion
Monroe Capital’s Ted Koenig didn’t name Apollo but didn’t need to. “Daily marks in illiquid private assets is an illusion of precision, not accuracy.” Fortress co-CEO Drew McKnight made the technical case. If a manager is marking to a leveraged loan index on spread, that can be done daily, “but it wouldn’t show you if the loan is at par unless you’re doing the actual analysis.” Koenig added that real marks need financial statements and inputs that don’t arrive daily.
Apollo’s response is that most of its $850bn in daily-priced assets are investment grade within its $40 trillion private credit vision, and that it has traded over $30bn of such loans. The SEC’s September reminder on fair value sits behind the whole debate.
Both sides are right about different things
Koenig is correct that a daily mark on a mid-market 1L with no financials since June is a model output, not a valuation. McKnight is correct that index-spread marking tells you about the market, not the borrower. Apollo’s point is that the alternative, a quarterly mark that moves 88 to 5 in nine months, is also a model output and a worse one. The defensible position is that daily spread-based marks capture rate and market moves while quarterly fundamentals capture the borrower, and investors should see both. The managers resisting daily marks mostly hold the kind of paper where the spread move is the part they’d rather not show.
4. The Discount Predicts the Markdown
A Georgetown and Ohio State paper found that listed BDC share prices carry information about future NAV changes before the NAV reflects them. A 10% lower price-to-NAV ratio is associated with a 0.6 to 1.9 percentage point higher probability of a subsequent NAV markdown, significant at one- and three-quarter horizons. Discounts also predict weaker net income and net investment income for up to four quarters. The effect is stronger at concentrated portfolios and at BDCs whose borrowers are held by fewer peers. The market is less able to predict specific non-accrual events or PIK changes.
Listed BDCs trade near 82% of NAV. Ares Capital is down 7% this year, FS KKR down 27%, Blue Owl Capital Corp down 18%.
The 82% is a forecast
Managers have spent the year describing listed BDC discounts as sentiment noise disconnected from portfolio quality. The paper says the discount is a forward indicator of NAV markdowns and earnings, and the relationship is strongest at exactly the concentrated, idiosyncratic portfolios where managers have the most discretion. At 82% of book, the sector-wide price implies the public market expects NAVs to fall. I’d read that as the same signal Saba is reading. Weinstein just paid about $17mm for 5.1% of TCPC, a fund down 27% this year with half its book already sold. He’s buying the discount because he thinks it’s partly wrong. The paper suggests it’s mostly right.
5. TCW Goes to a 30-Year Low on Credit
Bryan Whalen’s $160bn fixed income group at TCW has less exposure to investment-grade and high-yield bonds than at any point in three decades. Spreads don’t compensate for fundamental risk, he argued, and a blowup at the bottom of the market could spread. CCCs have traded at distressed levels since late September. “That is the weakest borrower, and they cry uncle first. Make no mistake about it, your single B borrower is going to feel it next, and then your double B borrowers.”
Whalen expects the consumer narrative to turn negative into year-end, sees AI infrastructure overbuilt, and holds agency MBS, ABS, and CMBS instead. His longer view is that a stretch of 10% high-yield defaults would be healthy for pricing.
The sequence runs through private credit
CCCs at 14% and distressed, then single-Bs, then BBs. Private credit’s borrower base sits squarely in the single-B and below bucket, with less covenant protection and more PIK than the public equivalents. Whalen isn’t talking about BDCs directly, but a $160bn allocator going to a 30-year low on corporate credit is the kind of institutional signal that eventually shows up in fund-finance terms and secondaries bids. His 10%-default-rate comment is worth filing. The industry’s base case is 6% and the bears say 9. Whalen is describing the clearing level as a feature.
6. Banks Review Their Australian Fund Lines
Goldman and UBS are reviewing the specialized financing they provide to Australian private credit managers, largely limiting new lending to the biggest platforms and those with modest property exposure. Existing commitments are mostly being maintained. About half of Australia’s A$200bn private credit market is tied to real estate. Warehouse and funding facilities to non-banks stood at A$98bn at end-2022, with later data not yet published.
The facilities span debt warehouses, NAV lending, and A/B note structures where the bank keeps the senior piece and leaves the fund the junior. Rates sit at a 15-year high with more hikes possible. “We are talking fundamentally about borrowing short and investing long in private markets,” said Zenith’s Dugald Higgins. “The pervasiveness of the linkages are always a little bit further than you thought.”
The quiet exit
Banks don’t pull facilities. They decline to renew them and tell the next fund that calls that the appetite is for larger platforms with less property. That’s what Goldman and UBS are doing in Australia, and it’s what HSBC did globally in the summer. Every gate in the US works because a revolver behind it stays open. Australia is where that assumption is being tested first, with half the market in real estate and a central bank still hiking. The sequence of Bathla, Metrics, KPMG, ASIC, and now the fund-finance banks has taken ten weeks. It’s a full cycle of what happens when the bank behind the gate decides the reward doesn’t cover the risk.
7. Europe Rebounds, Syndicated Takes the Big Ones
European direct lending volume hit its highest quarterly level of 2026 in the third quarter, with deals between €500mm and €1bn taking the largest share since mid-2023. Nine-month deal count and volume are still 16% below 2025. Direct lenders financed 20 buyouts in the quarter against six for the syndicated market, but syndicated quarterly volume reached €9.3bn, the highest since late 2021, and syndicated borrowers refinanced €2.6bn of private debt, the most in six quarters.
Average European direct lending spreads rose to 526 bps in 2026 after annual declines since 2022, while the median kept falling. France, around a fifth of European origination, is heading into a 2027 election with sovereign yields at quarter-century highs and a 2.8x downgrade-to-upgrade ratio at financial corporates.
The average rose because the tail widened
An average spread rising while the median falls means the market is splitting. Clean credits price tighter every year. Stressed and complex credits price wider, and there are enough of them to pull the average up. That’s a healthier market than 2024, when everything priced the same. The syndicated market taking €2.6bn of private debt and the largest buyouts in five years is the same dynamic as Mercer in the US. The borrowers who can leave are leaving. France adds a wrinkle, because the 20% of European origination that comes from there is about to go through an election where the leading candidates are promising debt cancellation.
Deals of Note
Volta Infrastructure - $5bn AI loan priced at S+650 and 95, yield above 11%; $1.3bn term loan plus $3.7bn chip loan for ~36,000 Nvidia GPUs; Anthropic signed six-year compute deal
Leidos/Analogic security units - Blackstone leading $1bn private credit financing for the combination
Omnes Education - French private school operator refinancing ~€700mm amid student unrest
Rodericks Dental Partners - Refinancing ~£360mm, likely with a PIK component
TCPC - Saba disclosed a 5.1% stake, paid about $17mm, in the ~$340mm market cap fund
IFC emerging-market vehicle - Tokio Marine as lead investor on a fund targeting up to $2bn
Bumble - Ithaca Holdings taking the dating app private for $1.3bn at a 138% premium
Shadowbox Studios - Silver Lake-backed Atlanta sound stage pivoting to a 651MW data center, facing a county with 40+ moratoriums
The Reality Check
The Blue Owl recording is a record of a firm that diagnosed its problem as a communications failure. Seven months later, the stock is at a low, bankers are pitching a sale, and the fund it wound down has a tenth of its remaining loans in default. The retail machine was the growth engine and the single point of failure, and no amount of KKR-in-2015 slides changes the fact that ~60% of OTIC came through one bank. The industry worked that out at Greenwich. Blue Owl is the case study.
Volta at S+650 and 95 is the number to carry forward on AI credit. A chip loan with a six-year Anthropic contract needed an 11% yield and call protection to clear, and the CLO base still didn’t want it. That’s the market pricing residual value on hardware in a way the private credit SPV deals from the spring did not. Every data-center and GPU financing marked in those books now has a visible comp 500 bps wider.
The BDC discount paper gives the 82%-of-NAV number a meaning managers won’t like. The public market isn’t confused. It’s forecasting markdowns, and the forecast is most accurate at the concentrated portfolios where the manager has the most room to mark generously. TCW cutting to a 30-year low, Goldman and UBS reviewing Australian fund lines, and Koenig calling daily marks an illusion are three responses to the same recognition. The marks are the argument now, and the people on the other side of it have started acting.