Private credit's first visible crack is actually the AI trade in disguise
Blue Owl's 5% redemption cap isn't a private-credit liquidity story. It's the first balance-sheet expression of SaaS-cannibalisation anxiety, and the industry's own investor letters are saying so in writing.
TL;DR
- Investors requested redemption of 41% of one US$6B Blue Owl fund and 22% of its US$36B flagship in Q1 2026. Both hit the 5% quarterly redemption cap.
- Blue Owl raised a US$400M bond at 6.5% to plug gaps. Treasury is now holding meetings with state insurance regulators about private-credit exposure.
- Industry-wide Q1 redemption requests across Apollo, Ares, Blackstone, Blue Owl and KKR: US$20.8 billion. The highest quarter since the 2022 regional-banking episode.
- The line that does the real work sits in Blue Owl's own investor letter: "heightened market concerns around AI-related disruption of portfolio-company revenue." The fund is redeeming because the loan book lent to companies the market now thinks AI will cannibalise.
- This is not a liquidity story dressed as a credit story. It is a SaaS-cannibalisation story dressed as a liquidity story.
What happened
On 14 April, Blue Owl Capital disclosed that two of its non-traded business-development company (BDC) vehicles had hit their quarterly 5% redemption caps. Investors in the smaller US$6 billion fund requested redemption of US$2.46 billion — 41% of NAV. Investors in the US$36 billion flagship requested US$7.92 billion — 22% of NAV.1
On 16 April, Blue Owl priced a US$400 million senior unsecured bond at 6.5% — roughly 180 basis points above comparable issuance from December 2025. The use-of-proceeds language reads "general corporate purposes including fund liquidity." On 18 April, the Wall Street Journal reported that Treasury officials had initiated meetings with state insurance regulators about private-credit exposure inside life-insurance balance sheets. Redemption requests for Q1 2026 across Apollo, Ares, Blackstone, Blue Owl and KKR totalled US$20.8 billion, per aggregated filings.
The sentence inside Blue Owl's investor letter that the firm's lawyers presumably tried to soften: "The pace of Q1 redemption requests reflects heightened market concerns around AI-related disruption of portfolio-company revenue, particularly in the software and business-services segments."
That is a private-credit fund telling its investors, in writing, that its loan book is being repriced because the equity market thinks the borrowers are going to be disrupted by AI.
What it actually means
Most of the coverage this week has been framed as a liquidity-and-rates story. Private credit got over-allocated, investors want out, higher-for-longer is biting. That frame is not wrong, but it is not the interesting part.
The interesting part is a loop that looks like this:
- Private-credit funds lent heavily to mid-market software and business-services companies between 2021 and 2024, usually at floating rates off SOFR with aggressive leverage multiples.
- Public software multiples compressed sharply in Q1 2026 on AI-cannibalisation fears — the SaaS universe down roughly 22% year-to-date against S&P +4%.
- Private software marks, which are set quarterly against public comparables under ASC 820 fair-value accounting, are now flowing through to private-credit fund NAVs.
- Investors notice their private-credit funds are down, notice that the funds lent to exactly the SaaS companies the market is downgrading, and request redemptions.
- Funds hit their 5% caps and raise debt at 180bps above December pricing to bridge.
The loop is not new. The feedback speed is. Redemption-cap gates are designed to compress the time between mark-downs and fund exits. They are not designed for the case where the underlying anxiety is structural rather than cyclical. AI cannibalisation is a structural hypothesis; waiting out a cycle does not resolve it.
The second thing to notice: insurance balance sheets hold roughly 28% of industry private-credit AUM as of end-2025 (AM Best data). State insurance commissioners are the regulators of record, and they have, until this week, been comfortable that private credit was a low-volatility fixed-income substitute. That comfort is the precondition for the asset class's growth from US$500 billion in 2015 to US$2.2 trillion in 2025. Treasury calling insurance regulators in to talk about it is the comfort being re-examined in real time.
Hype deconstruction — and also where the hype is under-done
"Private credit is having its 2008 moment." No. Leverage is lower, transparency is higher, and the gate mechanisms are doing exactly what they were designed to do. A 5% redemption cap that holds is a feature, not a bug. 2008 was marked by forced selling into illiquid markets; this episode is marked by controlled waitlists.
"This is a liquidity crisis." It isn't. Blue Owl raised US$400M at 6.5% in two days. Apollo drew on committed credit facilities without issue. Liquidity is expensive, not absent.
"Higher for longer is the story." Partially. Policy rates are certainly a contributing factor. But the Blue Owl investor letter specifically names AI disruption, not rates, as the redemption driver. If it were rates, private credit's own funding cost would be the problem — it isn't.
Where the hype is actually under-done: the mark-to-market contagion from public SaaS into private-credit NAVs. That mechanism is getting about 10% of the coverage it deserves. It is the transmission channel worth understanding.
Stakeholder landscape
| Stakeholder | Position | Pressure direction |
|---|---|---|
| Blue Owl, Apollo, Ares, KKR, Blackstone | Redemption-gated managers | Need to buy time without seeming to gate; raising expensive debt rather than force-selling loans |
| LP insurance companies | Captive balance-sheet holders | Regulators are asking questions; internal risk committees are requesting exposure reports |
| LP pensions and endowments | Redemption-requesters | Liquidity needs met for now; frustrated by waitlist; unlikely to recommit at existing rates |
| Mid-market software borrowers | Indirectly affected | Floating-rate resets into higher funding costs as fund WACC rises |
| Treasury / OCC / Fed | Monitors without jurisdiction | Private credit sits outside banking perimeter; can coordinate with state insurance regulators but cannot directly intervene |
| State insurance commissioners | Jurisdictional regulators | Being asked to raise their guard; historically deferential to fund managers |
| The AI trade itself | Structural cause | The same hypothesis that is lifting Nvidia is now being priced into borrower default probabilities |
Cross-layer implications
- Private credit fundraising. Q2 2026 fundraising will slow materially. Two managers have already quietly postponed fund closes. Expect a 20–30% decline in industry net inflows for H1 2026 against the 2025 run-rate.
- M&A financing. Private credit funds 60% of mid-market LBO debt. Waitlists absorb attention and dry powder. Expect sponsor-led M&A in the US$500M–US$2B range to slow by Q3.
- Insurance regulation. State commissioners now have political cover to tighten capital charges on private-credit holdings. NAIC is scheduled to revisit capital-charge methodology at its August 2026 meeting. The cover story for that meeting was drafted in April.
- Public SaaS. The feedback runs back the other way. Private-credit stress on SaaS borrowers means distressed secondary sales and cram-down restructurings, which mean equity wipes in public comparables. Expect two or three named mid-cap SaaS blow-ups in the next two quarters.
- Banks. A private-credit slowdown is mildly positive for direct lenders (JPMorgan, BofA, PNC). Expect trading-desk and DCM commentary in Q2 bank earnings to quietly celebrate.
What this means for you
If you are an LP. The redemption-cap is not a one-quarter event. Waitlists of this size have historically taken four to seven quarters to clear. If you need the capital inside 12 months, tender at secondary pricing (currently 89–92 cents on the dollar for Blue Owl BDC interests) rather than queue. If you can hold 24 months, the waitlist clears at NAV.
If you are a GP of a private-credit fund. Run the SaaS-exposure test. If more than 25% of your loan book is in software or business-services, expect LP questions in every Q2 diligence call. Pre-empt with an exposure breakdown rather than answer it reactively.
If you are a borrower. Expect amendment-and-extend conversations to start inside 90 days. Covenant-lite structures from 2022–23 are the ones most exposed; your sponsor will be calling your fund first. Open the line.
If you are a CFO at a mid-market software company. Your equity sponsor is now a risk factor in your cap table, not just a source of patience. Revenue-cannibalisation tests from AI are about to be requested quarterly, not annually. Build the deck before you are asked for it.
If you are a regulator. The transmission mechanism — public-market SaaS anxiety → private-market software marks → private-credit NAV → insurance balance sheet — is the one to instrument. Existing Form PF data captures the first three nodes. The fourth node is captured in state-insurance filings that are not federally aggregated. That gap is the one to close.
Uncertainty ledger
- Waitlist visibility. Funds disclose redemption requests quarterly, but the waitlist stack is not consistently reported. The true industry waitlist number is estimated at US$8–14B; the range reflects disclosure heterogeneity.
- Mark-to-market methodology. ASC 820 "fair value" for private loans allows managers discretion in selecting comparable-company universes. A rewrite of the universe can move NAV by 3–6% without any change in underlying credit.
- AI-cannibalisation evidence. The hypothesis is being priced in equity and spreads faster than it is being observed in revenue. Q2 2026 SaaS earnings will be the first test of whether the hypothesis has landed in actuals.
- Regulatory action. Treasury meetings with state insurance regulators are not a regulatory action; they are a prelude. Whether they become action depends on whether Q2 redemptions accelerate.
Bottom Line
Blue Owl did not have a liquidity problem this week. It had the first visible case of AI-cannibalisation anxiety moving from equity markets into credit marks and from credit marks into fund flows, and its own investor letter admits this. The gate mechanics are holding, which is what gates are for. What they cannot do is resolve the structural question — whether the borrowers are going to get revenue-disrupted — because that question is not cyclical. Anyone who tells you this is a higher-for-longer story is telling you what their compliance department approved. The interesting story is already in Blue Owl's investor letter, in the sentence the firm's lawyers let through.
Written in the tradition of — M.
Sources
- Tier 1: Blue Owl Q1 2026 investor letter; SEC BDC filings (Apollo, Ares, Blackstone, Blue Owl, KKR); Wall Street Journal; Financial Times; Reuters
- Tier 2: Bloomberg; AM Best industry data; Preqin private-credit AUM series
