Private Credit’s Redemption Test: What the 2025–2026 Wave Means for U.S. Direct Lenders, Venture Capital, and Return-Seeking Investors
Abstract
The U.S. private credit market has entered its first broad liquidity confidence stress test since it became a mainstream asset class for both institutions and private wealth. The immediate trigger has not been a generalized collapse in loan performance, but a sharp rise in redemption requests in retail- and semi-liquid structures, especially non-traded BDCs and interval funds after a sequence of shocks: the bankruptcies of First Brands and Tricolor in 2025, growing scepticism about private valuations, and renewed concern that AI disruption could impair software borrowers, one of the sector’s most heavily financed end markets.
The recent wave in February 2026 has been led by vehicles associated with Blackstone, BlackRock/HPS, Blue Owl, and Cliffwater, while Ares and other large platforms have reported elevated redemption pressure even where formal gates were not yet the headline. The episode matters because it reveals a structural tension at the heart of modern private credit: firms raised vast pools of capital by packaging illiquid loans into vehicles marketed as income-producing, low-volatility, semi-liquid alternatives for private wealth, yet the underlying assets remain loans that cannot be sold instantly without potentially impairing value.
For venture capital, the consequences are unlikely to be uniform. Instead, this is producing a barbell outcome: debt remains abundant for asset-heavy or strategically favoured themes such as AI infrastructure, while lenders are becoming materially more selective toward legacy SaaS, business services, and other sectors now deemed vulnerable to AI-driven margin compression. That, in turn, should affect venture debt availability, late-stage valuation support, and the pace of sponsor-led exits globally.
1. The Key Players and the Anatomy of the Recent Redemption Wave
The most visible recent flashpoint was Blackstone’s BCRED, an approximately $82 billion flagship private credit fund, where first-quarter 2026 redemption requests reached 7.9% of NAV, prompting Blackstone to raise its usual redemption limit from 5% to 7% and to contribute $400 million of firm and employee capital so that all requests could be met. Reuters described this as BCRED’s first quarter of net outflows, with approximately $3.7 billion redeemed and net withdrawals of roughly $1.7 billion after new commitments.
A second major event came from BlackRock’s HPS Corporate Lending Fund (HLEND), a roughly $26 billion non-traded BDC, which disclosed that investors sought to redeem 9.3% of shares in Q1 2026; management paid only the standard 5% repurchase amount, or roughly $620 million, thereby formally capping withdrawals. Bloomberg characterized this as one of the clearest major examples of gating in the current cycle.
A third signal was Cliffwater Corporate Lending Fund, a roughly $33 billion interval fund, where investors reportedly requested redemption of about 14% of shares in Q1 2026 and Cliffwater responded by honouring only 7%, describing that amount as the “regulatory maximum.” This was one of the largest withdrawals attempts yet reported in a retail-facing private credit vehicle.
The most controversial restructuring of liquidity came from Blue Owl. In February 2026, Blue Owl said it would sell $1.4 billion of direct lending assets across three funds and permanently halt regular quarterly redemptions at OBDC II, replacing tender-based exits with episodic return-of-capital distributions. Reuters reported that OBDC II had faced persistent redemption pressure, while Blue Owl’s broader credit complex had already been under scrutiny because of software exposure and liquidity management concerns.
The broader pattern was already visible in late 2025. Bloomberg reported data cited by Wealth Management showed that investors in BDCs larger than $1 billion sought to withdraw more than $2.9 billion in Q4 2025, up roughly 200% from the prior period, with elevated requests at Blackstone, Blue Owl, and Ares. Fitch then confirmed that average redemptions across perpetually non-traded BDCs rose to 4.5% of NAV in 4Q25 from 1.6% in 3Q25, with several funds funding tenders above the usual 5% cap.
This is why the best description of the economic wave is not “systemic insolvency,” but a concentrated redemption wave in semi-liquid wrappers. Bloomberg explicitly noted that many executives saw BlackRock’s move as industry cover for aligning fund liquidity with asset liquidity, while Fitch emphasized that liquidity cushions are still generally adequate, even if sustained tenders above 5% would pressure credit profiles.
2. How These Private Credit Players Operate and How They Raised AUM So Quickly
The fastest-growing U.S. private credit firms scaled by exploiting a structural gap left by banks after post-2008 regulation curtailed balance-sheet lending to middle-market and sponsor-backed borrowers. In effect, private credit firms became non-bank balance sheets: they originate floating-rate senior secured loans, often hold them to maturity, and sell the strategy as a blend of yield, downside protection, and low mark-to-market volatility. Reuters summarized the sector as a market that has ballooned to roughly $2 trillion, while S&P Global placed global private credit above $1.7 trillion in late 2025.
The biggest firms also industrialized origination. Apollo reported that in 2025 it generated more than $300 billion of originations and more than $225 billion of inflows, with total AUM of roughly $938 billion and credit AUM of about $749.2 billion, making credit the centre of its business model. Apollo ’s competitive edge has been the fusion of direct origination, insurance capital, and perpetual-capital vehicles that allow the firm to raise assets continuously rather than in traditional closed-end vintages.
Ares has used a similar playbook at scale. As of December 31, 2025, Ares reported approximately $623 billion of AUM and emphasized its global credit platform as the core of its franchise. Blue Owl likewise reported roughly $307.4 billion of AUM, including $157.8 billion in credit AUM and $115 billion in direct lending, showing how quickly a distribution-first platform can reach scale when it combines institutional origination with private-wealth access.
Just as important as origination was distribution. Private credit managers moved beyond institutional LPs and aggressively targeted financial advisors, high-net-worth investors, family offices, and eventually retirement-linked channels, primarily through non-traded BDCs, interval funds, evergreen funds, and perpetual-capital structures. With Intelligence estimated that evergreen funds had surpassed $500 billion in AUM by 2025, while Fitch listed a broad universe of perpetually non-traded BDCs spanning platforms such as Apollo, Ares, Bain, Barings, BlackRock, Blackstone, Blue Owl, Carlyle, Goldman Sachs, Golub, HPS, Monroe, North Haven, Oaktree, and Antares.
Marketing mattered. These products were positioned not as speculative credit, but as institutional-quality direct lending with regular income, often emphasizing senior secured positioning, floating rates, diversification, lower volatility than public high yield, and access to an otherwise exclusive institutional asset class. BCRED’s own shareholder materials stress its “strength in scale,” senior-secured portfolio, and distribution profile, while Apollo’s current response to market scepticism has been to promise more frequent NAV reporting and ultimately daily valuations to reinforce a transparency narrative.
The essential point is private credit did not gather assets so quickly simply because institutions loved direct lending economics. It grew because managers productized illiquidity, wrapped it in smoother-return vehicles, and sold it through wealth channels that were previously underpenetrated by alternatives. The redemption wave is therefore as much a distribution-model test as it is a credit test.
3. Why the Redemption Wave Matters for Global Venture Capital: Two Specific Transmission Channels
3.1 Venture debt will not disappear but it will become more selective, especially for software
One immediate transmission channel to venture capital is through the repricing of software risk. Reuters reported in February 2026 that software companies were delaying debt deals as lenders demanded higher spreads and subjected borrowers to tougher scrutiny because AI may erode traditional software business models. UBS, cited by Pitch Book, estimated that 25%–35% of private credit portfolios face elevated AI disruption risk, with particularly acute exposure in technology and business services.
That matters because venture debt has simultaneously become larger and more important. Pitch Book-NVCA data showed U.S. venture debt reaching a record $62.4 billion in 2025, even as deal count fell, implying larger average facilities and deeper participation by major private lenders. In other words, debt capital is still abundant but it is being redistributed toward borrowers viewed as more defensible.
Specific example: a late-stage vertical SaaS company pursuing a bridge loan between equity rounds now faces a meaningfully different market than it did in 2024. If the company depends on seat-based pricing, weak switching costs, or undifferentiated workflow software, lenders may reduce facility size, widen spreads, demand stronger covenants, or insist on additional equity support. That raises dilution risk for founders and existing VCs because the firm may need to raise equity sooner, and potentially at a lower valuation.
3.2 Debt will rotate toward AI infrastructure and asset-backed themes, widening dispersion in VC outcomes
A second transmission channel is capital rotation, not merely capital withdrawal. Pitch Book reported that AI infrastructure and larger, asset-heavy startups are pulling in very large debt facilities, citing Crusoe’s $750 million debt round from Brookfield in June 2025 and xAI’s $5 billion loan from Morgan Stanley in July 2025. KPMG’s Venture Pulse further showed that global VC investment exceeded $512 billion in 2025 and remained highly concentrated in AI.
Specific example: an AI infrastructure, data-centre, power, or compute-adjacent startup can still secure debt because its assets, contracted cash flows, and strategic relevance fit what lenders increasingly prefer. By contrast, a generalist software startup may find that both venture debt and sponsor-backed takeout financing are less available, even if headline VC funding remains strong at the top end of AI. The effect is not a collapse in venture capital, but a much steeper dispersion curve between debt-friendly and debt-unfriendly sub-sectors.
A third-order effect also matters: Bloomberg warned that meeting large redemption requests can divert capital from new deals, reducing risk appetite for fresh financings. If that persists, growth-equity and buyout firms that often provide exit paths for VC-backed companies may slow acquisitions, especially in sectors where software multiples are already under pressure.
4. Which Companies and Vehicles Are Most Exposed and What Is the Rational Course of Action?
The directly exposed fund platforms include the most important retail-facing and semi-liquid private credit vehicles in the market: BCRED (Blackstone), HLEND and BDEBT (BlackRock/HPS), OBDC II, OCIC, and OTIC (Blue Owl), ASIF (Ares), ADS (Apollo Debt Solutions), and the broader perpetually non-traded BDC universe tracked by Fitch, including Antares Strategic Credit Fund, Bain Capital Private Credit, Barings Private Credit Corp., Carlyle Credit Solutions, Goldman Sachs Private Credit Corp., Golub Capital Private Credit Fund, Monroe Capital Income Plus, North Haven Private Income Fund, and Oaktree Strategic Credit Fund.
The listed BDC segment also signals where public investors believe pressure may be greatest. Reuters reported that by March 2026 many major BDCs traded at steep discounts to NAV, including FS KKR Capital Corp., Blue Owl Technology Finance, Prospect Capital, Carlyle Secured Lending, Blackstone Secured Lending, and even Ares Capital at a smaller discount. Discounts do not prove loan impairment, but they did show the market is questioning marks, sector exposure, and recovery assumptions.
On the borrower side, the most exposed operating companies are those with some combination of software concentration, weak cash conversion, aggressive add-on acquisitions, near-term maturities, covenant resets, or dependence on valuation support rather than free cash flow. Fitch highlighted software as a key exposure area for BDCs and alternative managers; Reuters noted that software companies were already pausing debt deals; and JPMorgan’s recent markdowns specifically targeted software-linked collateral in some private credit portfolios.
A. Course of action for affected companies (operating businesses and portfolio companies)
A prudent course of action for operating companies is not to wait for a maturity wall or covenant tripwire. The first priority is to secure 18 to 24 months of liquidity under a downside case, not a base case. That means stress-testing revenue assumptions, reducing burn, prioritizing cash conversion over ARR optics, and approaching lenders early for amend-and-extend discussions before the borrower becomes a forced negotiator. The current market is penalizing surprise, not just weakness.
The second priority is capital-stack diversification. Companies dependent on a single direct lender or a single venture debt provider should explore multiple channels for instance, bank debt where available, structured preferred equity, asset-backed lines, receivables financing, or sponsor support because the lesson of the current episode is that liquidity can evaporate more quickly in a concentrated funding model than management teams expect.
The third priority is re-underwriting the business model through an AI lens. Firms whose value proposition is not mission-critical, data-rich, or embedded in regulated workflows should assume higher refinancing spreads and lower leverage tolerance. Fitch has explicitly argued that not all software is equally vulnerable and that lenders are differentiating between critical systems-of-record and more replaceable tools. Companies should therefore frame their financing story around defensibility, switching costs, proprietary data, and cost-to-replace, and not only historic growth.
B. A non-personalized framework for investors who seek “high alpha” and aspire to >15% IRR
For investors seeking returns above 15%, the current environment does not reward broad beta exposure to generic direct lending. Instead, the market is likely to favour dispersion, complexity, and seniority with control. In practical terms, that usually means strategies such as structured preferred equity, rescue financing, continuation-vehicle secondaries, capital-solutions transactions, selective NAV-backed financing, and specialty finance or asset-backed segments where collateral and cash-flow visibility are stronger than in generalized sponsor software loans. These strategies are inherently more complex and higher risk, which is precisely why the return target is possible.
Equally important, the underwriting bar should now be higher than it was during the easy asset accumulation phase of private credit. A rational institutional framework would demand: lower entry leverage, stronger covenants, more conservative EBITDA adjustments, explicit downside collateral analysis, sector-level AI disruption screens, and sponsor behaviour analysis. Reuters, Fitch, and JPMorgan-related reporting all point to the same lesson: the market is no longer willing to assume that private marks, software cash flows, and refinancing conditions will remain benign.
The final point is philosophical: a 15%+ IRR in the present climate, in 1st quarter in 2026, is most plausibly earned by being the provider of scarce, patient, highly structured capital when others are liquidity constrained and not by owning crowded, semi-liquid yield products marketed as cash substitutes. The redemption wave has made that distinction unmistakable.
Conclusion
The current U.S. private credit redemption wave is not, in itself, proof that the asset class is broken. It is, however, proof that the most successful managers built enormous franchises on a promise that required careful choreography. In other words, illiquid loans funded by investors who increasingly expected periodic liquidity. Once valuation doubts, software risk, and a handful of credit blowups collided, redemption pressure rose exactly where that choreography was most delicate, especially in the retail and semi-liquid wrappers that powered asset growth.
For venture capital, this means a more selective debt environment, not a generalized funding drought. Late-stage software and marginal SaaS models are likely to face tighter leverage, higher spreads, and more down-round risk, while AI infrastructure, data-centre, and asset-backed themes may continue to draw both debt and equity capital at scale. In other words, private credit stress is likely to increase dispersion in venture outcomes, not eliminate venture financing.
For companies and investors alike, the strategic imperative is the same: move from asset gathering to underwriting discipline. In the next phase of the cycle, alpha is less likely to come from merely participating in private credit and more likely to come from knowing which parts of the private credit complex deserve capital, which deserve patience, and which deserve a hard reset of assumptions
