Newsletter · · Ashutosh Agarwal

Banks Retreat From Private Credit as Managers Call It Noise - The Private Credit Boom (and Cracks) - Week of July 15, 2026

Financials newsletter for the week of July 15, 2026. Big banks led by HSBC began pulling back the leverage they lend to private-credit funds and marking down collateral, and for the first time in weeks the managers answered on the record, with Blackstone calling new-GFC talk irresponsible and Apollo's economist sizing the real crack at roughly $500 billion of software loans due in 2028–29.

The Private Credit Boom (and Cracks)

Week of July 15, 2026: Banks Retreat From Private Credit as Managers Call It Noise

TL;DR

  • The banks are backing away. HSBC told some private credit funds it won't renew their credit lines and is pulling the leverage it lends them, and JPMorgan, Goldman Sachs and Barclays have reportedly been marking down loans those funds pledged as collateral. When the people who fund the machine start repricing the risk, that's worth more than any marketing deck.
  • The managers pushed back, hard, and on the record. After a quiet week where the only voice was a bear, this week the operators showed up. Blackstone's private-wealth chief called the "new GFC" talk "irresponsible and just wrong," and Apollo's chief economist laid out data showing defaults are actually falling, with one glaring exception.
  • That exception is software. Apollo's Torsten Slok says roughly $500 billion of the ~$2 trillion private-credit market is software loans, many of them hitting a refinancing "wall" in 2028–29. That's the real crack, and, he argues, a contained one.

What's New

1. HSBC heads for the exit, and it's not alone. The single biggest development this week came from Jeff Snider, a market commentator, on Eurodollar University (July 9). Citing Financial Times reporting, he flagged that HSBC "has informed some of its private credit borrowers it's not going to renew their credit facilities" and is "pulling some back leverage away from risky private credit funds," after "a series of high profile corporate bankruptcies exposed... weaker underwriting standards across at least parts of the industry."

A quick plain-English detour, because this is the whole story: back leverage is money a bank lends a private credit fund against the fund's own loan book. The fund makes loans to companies, then borrows against those loans to make even more loans and juice returns. HSBC's message, in effect: the collateral you're posting isn't worth what you say it is, and the return we get for financing you no longer covers the risk. Snider also cited Bloomberg reporting that big banks, "JPMorgan, Goldman Sachs, and Barclays, in addition to HSBC," have been "raising rates on leverage" and "marking down individual loans that are posted as collateral," forcing fund managers to swap other assets into their collateral pools. Why it matters: this is the funding layer voting with its balance sheet before any headline default shows up. It reads straight through to Blue Owl, Blackstone and Apollo.

2. Blackstone's operator answers the "new GFC" crowd directly. On Alt Goes Mainstream (July 9), Farhad Karim, COO of Blackstone Private Wealth, didn't dodge the bear narrative, he took a swing at it: "You read many of the articles, you think the whole private credit or the whole private... market is melting down. The system is not melting down. And you hear strong language around, are we in the new GFC? We're not. I mean, in fact, some of this commentary I find pretty irresponsible and just wrong." Why it matters: it's the manager's on-the-record rebuttal to exactly the story making the rounds this week, from the firm with the most retail-wealth money at stake.

3. The retail-and-insurance money keeps pouring in. Same conversation, the number that anchors the bull case: Blackstone's private-wealth business is now at "302 billion of AUM" (assets under management, the pile of client money it runs), up from about $250 billion 18 months ago, and management's stated ambition is "a trillion of AUM within the Wealth Channel." Karim says Blackstone has "reached 18,700 advisors," is building its biggest overseas push in Japan (eventually "50 people on the ground") followed by Canada, and increasingly sells one-ticket "multi strategy" funds plus a public-private hybrid built with Vanguard and Wellington. He frames three money pools, "individual money, institutional money, insurance money," working "in a particular synergistic fashion." Why it matters: this permanent-capital flywheel is the reason bulls think private credit can grow through a rough patch, the money coming in the front door isn't the hot money that runs at the first bad headline.

4. Apollo's economist names the crack, and sizes it. On The Real Eisman Playbook (July 13), Torsten Slok, Apollo's chief economist, sat with investor Steve Eisman and made the most data-grounded argument of the week. His headline: "if you look at default rates in loans and in high yield, they have actually been going down for the last 12 months," along with distressed debt swaps and other rescue restructurings, "so let's agree at the highest level, credit is actually getting better." The exception: software. Software borrowers carry lots of debt and thin ability to service it (a weak "coverage ratio," earnings divided by interest cost). Slok sizes it precisely: private credit is "a $2 trillion market" and "$500 billion of private credit is software that was originated in the last six, seven years." Those loans now trade cheap enough to yield "12%... moving up towards 12.5%." The timing risk: software loans written in "2021 and 22" with "a seven-year maturity" mean "in 2028 and 29, we are running into the maturity wall," the point where the loan comes due and has to be refinanced. Why it matters: it turns a vague fear ("private credit is cracking") into a specific, dateable one (software, 2028–29), and it comes from inside one of the biggest managers.

5. An operator confirms the growth themes, and the danger. On the Money Maze Podcast (July 9), Sixth Street Co-CIO and Co-President Julian Salisbury said "capital solutions providers is a pretty rich theme for us right now," companies that locked in cheap debt in 2020–21 now can't refinance normally, so firms like his step in to pay down debt, return cash to investors, or fund growth. He also flagged rising demand from insurers for asset-based finance (lending against pools of assets like loans or receivables) as a way to "earn a liquidity premium... above what they can earn in the public bond markets." But he added the warning bulls should sit with: if a manager gets "ill disciplined about the type of money you raise... you find yourself under pressure to invest more capital than is prudent," turning "from being investment firms to being sales led organizations." Why it matters: it's an insider validating the ABF-and-insurance growth story while naming the exact mechanism, asset-gathering outrunning underwriting, that the bears say is already happening.

The Debate

This week finally gave us both sides in full voice.

The bear case (cycle turning): The people with the clearest view, the banks that finance the funds and see the actual loan books, are quietly pulling back. HSBC won't renew lines; JPMorgan, Goldman and Barclays are marking down collateral. Bankruptcies have already exposed sloppy underwriting at "at least parts of the industry." Meanwhile, the marks that make private credit look smooth are model-based, not real trades, and only hold up "as long as everyone believes in them." When banks stop believing, funds have to post more collateral, sell loans, or accept worse terms, all of which hurt returns and liquidity at the same time.

The bull case (idiosyncratic noise in a growing asset class): Defaults, distressed swaps and rescue restructurings are all falling, per Apollo's own economist. The real trouble is concentrated in one sector, software, ~$500 billion of a ~$2 trillion market, and even that is a fraction of a $33 trillion economy. Crucially, it isn't sitting on dangerously leveraged balance sheets: business development companies (BDCs, the publicly traded funds that make these loans) are capped by law at 2:1 leverage, versus banks that ran at "20, 30... 40" times in 2008. So there's no forced-selling machine to turn losses into a crisis. And the money funding the industry is increasingly sticky retail and insurance capital, not flighty hot money.

The honest middle: both can be true for a while. Snider's own framing is that "public markets can stay calm long after private funding conditions begin to deteriorate." Slok would agree the software vintages are a genuine problem, he just thinks it's a contained one that bites in 2028–29, not a system-wide 2008.

Pull-quote of the week, from Snider, capturing the split-screen perfectly:

"While the Blue Owls and the John Grays are running around telling you their numbers all work and their portfolios are just fine, the banks who are looking right at those same portfolios are getting the hell away from them. Take the inside information every time."

And the operators' answer, from Blackstone's Karim:

"The system is not melting down... some of this commentary I find pretty irresponsible and just wrong."

You now have the whole argument in two sentences. Pick your side.

Stocks in Play

Blackstone (BX / BDC: BXSL)

  • Bull: The wealth machine is compounding, $302B and climbing toward a $1T target, 18,700 advisors, sticky insurance and retail money, and the scale to "speak for" massive AI data-center deals without syndicating them out. Management is publicly, confidently rebutting the crisis narrative.
  • Bear: Snider named "the John Grays" (Blackstone President Jon Gray) specifically as the people telling you everything's fine while banks retreat. If collateral marks and redemptions bite, the firm most levered to the retail-wealth story has the most narrative to lose.
  • Next catalyst: Q2 earnings and any color on BXSL non-accruals, fund redemption/gating activity, and wealth-channel inflows.

Apollo (APO)

  • Bull: Its own chief economist is making the most credible "credit is actually getting better" case on data, and Apollo is reportedly "moving up in credit quality" and holding more cash, positioning for choppier conditions rather than getting caught out. Apollo funds are still paying "8%, 9%, 10%."
  • Bear: Last week's coverage flagged Apollo among managers that "put limits on the amount of money investors can take out of these funds." Redemption gates are the classic tell that liquidity is tightening.
  • Next catalyst: Earnings and any update on redemption limits, spread-lending activity, and insurance (Athene) deployment.

Blue Owl (OWL / BDC: OBDC)

  • Bear: Called out by name ("the Blue Owls") as a manager whose public "portfolios are fine" message the banks apparently don't share. Heavy software/tech-lending exposure is precisely the sector Slok flags.
  • Bull: No operator from Blue Owl was on any podcast this week to answer; the read-through from Slok's data (broad credit improving, software the outlier) cuts both ways depending on OBDC's software mix.
  • Next catalyst: OBDC's software exposure disclosure, PIK share (paid-in-kind, interest paid with more debt rather than cash, a stress signal), and non-accruals at next report.

Ares (ARES / BDC: ARCC)

  • Note: No Ares operator or ARCC-specific commentary surfaced on podcasts this week. Ares appeared only in last week's mention of redemption limits alongside Apollo. Watching for the same software-exposure and non-accrual questions at earnings.

HSBC, JPMorgan, Goldman Sachs, Barclays

  • Read-through, not a call: These are the banks reportedly de-risking. If more banks follow HSBC in cutting back-leverage lines, it raises funding costs across the private-credit industry, a headwind for the managers and a possible tell that bank credit officers see something the public marks don't.

Read-throughs

  • BDCs (ARCC, BXSL, OBDC): The structural shield bulls keep citing, 2:1 legal leverage cap, is real and is the single best argument against a 2008 rerun. But the software concentration Slok flags lives partly inside these portfolios. Watch non-accruals, PIK share, and NAV marks next earnings; watch whether rising bank leverage costs squeeze BDC economics.
  • Insurance balance-sheet partners: Both the bull (Blackstone's "insurance money" leg) and the operator (Sixth Street's ABF demand from insurers chasing a yield premium) point the same way, insurance permanent capital is still a growth engine. The risk is that insurers are reaching for the same complexity premium right as banks decide the risk isn't worth it.
  • Regional / lending banks: The bigger picture Slok drew is a post-Dodd-Frank world where "banks were essentially asked to do less, and the market was asked to do more." Private credit taking share from bank lending is a multi-year structural loser for bank loan books, even if this week's HSBC story is banks stepping back from financing private credit rather than competing with it.
  • Syndicated loans / CLOs: No direct CLO coverage on podcasts this week. The relevant tell is Snider's "split screen": publicly traded high-yield spreads remain "exceptionally tight" even as private funding tightens, he attributes the calm to mechanical passive/60-40 fund flows, not fundamentals. If private stress spills into public credit, tight spreads are the thing that snaps.
  • Data-center / ABF borrowers: The one theme both sides are bullish on. Blackstone can "speak for" massive AI data-center deals solo; Sixth Street backed an early-stage data-center platform that later "naturally transitioned to a lower cost of capital." Asset-based finance and data-center lending remain where the growth capital wants to go, the question is discipline, per Salisbury's warning.

What Changed vs Last Week

Last week (July 1–8) was thin, one qualifying episode, and it was a pure bear call: analyst-commentator Chris Whalen arguing public BDCs were "unprofitable now" and "not a time to be investing," with no manager or operator voice to answer him.

This week the picture filled out on both sides:

  • The operators finally showed up. Last week's biggest gap, "no operator/manager voice," closed. Blackstone (Karim), Apollo (Slok), and Sixth Street (Salisbury) all spoke, and two of them directly rebutted the bear narrative.
  • The bear case escalated from opinion to sourced action. Last week was a pundit's view; this week is banks actually pulling credit lines and marking down collateral (HSBC, JPMorgan, Goldman, Barclays, via FT and Bloomberg reporting). That's a harder data point than "BDCs look unprofitable."
  • The redemption-gating theme was corroborated. Last week noted Apollo and Ares limiting investor withdrawals; this week Snider independently flagged rising redemptions and Apollo "moving up in credit quality... holding more cash."
  • Software / AI-adjacent credit stress got quantified. Last week gestured at AI/data-center spreads widening; this week Slok put numbers on it, $500B of software in a $2T market, 12%+ yields, a 2028–29 maturity wall.
  • Still missing, two weeks running: hard fee-related-earnings and dry-powder figures for the big managers, BDC-specific non-accrual/PIK data, and a clean new-loan-spread-vs-syndicated-loan comparison. Those live in earnings and filings, not on podcasts, and earnings season is the near-term fix.