U.S. Private Credit Risks: Small-Scale Subprime Crisis?

Ask AI · How could AI technology breakthroughs ignite hidden risks in private credit?

By: Ze Ping Macro Team

For a long time, the U.S. private credit market has relied on opaque valuations, liquidity mismatches, and PIK structures to maintain calm on the surface, and AI’s disruption to the software industry has punctured that layer of sturdy-looking armor.

In early April, a fund under Blue Owl faced a 40% redemption request rate; giants such as Blackstone and BlackRock moved to restrict withdrawals one after another. In a rare warning, Buffett issued an alert, and the U.S. Treasury urgently convened a regulatory meeting.

What caused this U.S. private credit turmoil? Could it evolve into a second “subprime crisis”?

1 Event: A wave of redemptions spreads across U.S. private credit funds, prompting emergency action by regulators

Over the past decade, private credit has expanded rapidly in global financial markets in the role of “bank substitute.” Private credit is loan and financing provided to enterprises by non-bank institutions such as private credit funds and Business Development Companies (BDCs), filling the credit supply gap left by banks as tighter regulation reduces lending supply. According to IMF data, the global private credit market has already exceeded $2 trillion in size, and roughly three-quarters are concentrated in the United States. The market share is approaching that of syndicated loans and high-yield bonds.

Since 2026, private credit funds have faced successive redemption waves and liquidity pressures, starting to restrict redemptions; an industry-wide trust crisis has spread. In February 2026, Blue Owl, an alternative investment company, announced the permanent closure of the redemption window for a $1.6 billion private credit fund under its umbrella. In early April, two private credit funds under Blue Owl saw large-scale redemptions in the first quarter: for the OCIC fund, redemption requests accounted for 21.9% of the issued shares; for the OTIC fund, as high as 40.7%. Combined, the total scale was about $5.4 billion. Under pressure from massive redemption demand, Blue Owl set an upper limit on the actual redemption ratio at 5%, and attributed the abnormal increase in redemption requests to “growing market concerns that AI may impact software companies.” Limiting redemptions is not an isolated case—recently, industry heavyweights such as Apollo, BlackRock, and KKR have all faced redemption restrictions and liquidity pressures, with panic sentiment continuing to spread. Investors’ confidence in private credit funds has been shaken. According to BIS data, BDCs (Business Development Companies) with higher SaaS exposure saw even deeper declines.

Banks tighten financing, Buffett warns, and risk signals keep being released. On March 31, Reuters reported that several U.S. banks had increased the financing costs for private credit funds, and that spreads on some BDC-related financing rose to about 2 percentage points above SOFR. JPMorgan lowered the valuation of related collateral. On March 30, Federal Reserve Chair Powell also clearly stated that he is closely monitoring risk spillover from private credit; while he does not yet believe it poses a threat to the entire financial system, he has explicitly warned that losses will appear in the relevant areas. In an interview with CNBC, Buffett pointed out that as links between the banking system and non-bank financial institutions (such as private credit funds) have become increasingly tight, signs of fragility have begun to appear in the financial system. Berkshire currently holds more than $350 billion in cash, and Buffett has clearly stated that the current market has not met his threshold for action.

The U.S. Congress issues a warning letter, and the U.S. Treasury convenes regulators to discuss private credit risk. As multiple top asset management institutions restricted withdrawals one after another, on April 1 the U.S. Treasury announced that from April through early May it will convene U.S. and non-U.S. insurance regulators, focusing on recent developments in the private credit market, emerging risks, risk-control practices, and industry outlook. The discussions will center on four key topics: fund leverage, private ratings, offshore reinsurance, and investment liquidity. It also made clear that this meeting will be the beginning, and that a longer-term communication mechanism will be established afterward. The House Financial Services Committee has sent requests for information to multiple major institutions, including Blackstone, Ares, Apollo, BlackRock, Blue Owl, and others. The scope covers sales practices, leverage levels, fee structures, incentive mechanisms, audit arrangements, risk management, and potential economic vulnerabilities.

2 Cause: Risk-hidden mechanisms fail; core assets are repriced

The outbreak of this crisis is not accidental, but the result of multiple structural factors accumulating over the long term.

First, the private credit industry has long relied on valuation inertia and the extension of cash flows to keep things appearing calm, while internal vulnerabilities continue to build. After the 2008 financial crisis, banks exited parts of higher-risk lending under strong regulation. Private credit stepped in to fill the financing gap, masked price volatility through non-public transactions, and then created the impression of “high yield, low volatility, and periodic exits” via product design aimed at wealth management and the retail end. By 2025, private credit has accounted for about 30% of the U.S. leveraged financing market. But this apparent stability is not because the assets themselves are safer; rather, it comes from opaque valuations, liquidity mismatches, and the shifting of stress to later periods.

Second, the overuse of PIK structures uses “debt accumulation” to conceal real defaults. In the past two years, private credit institutions have also heavily used PIK (payment in kind) structures, allowing borrowers to temporarily not pay cash interest and instead maintain turnover through the accumulation of debt. By the end of 2025, in private credit contracts corresponding to software borrowers, more than 20% already included PIK options—doubling over three years. These arrangements can suppress reported delinquency rates in the short term, but they postpone the exposure of real risk, and instead increase the likelihood of future defaults.

Third, AI disruption driving a repricing of software assets punctures this layer of stability armor. Software enterprises—the type of core assets that private credit relies on most for its growth narrative—are being upended by AI technology and are being repriced first. Over the past few months, events such as First Brands and Tricolor have already raised market doubts about underlying lending standards, while the breakthrough development of AI technology has further shaken heavily concentrated software holdings across the industry. Blue Owl attributed the surge in redemptions in this round to market concerns about AI’s impact on software companies. About 8% of its managed assets are allocated to software. Barclays and Morgan Stanley have also warned that concentrated exposure to the software sector will push up future default rates. Morgan Stanley estimates that from the second half of 2026 to the first half of 2027, the industry’s annualized default rate could rise to 8%. Once the growth narrative that supports industry valuations fails, credit quality will worsen, financing will tighten, and redemptions will accelerate. Previously hidden risks become concentrated and manifest explicitly, turning into the direct trigger for this round of turmoil.

In summary, the U.S. financial system has not truly absorbed high-risk credit; instead, it has transmitted it to more investors that appear, at least on the surface, to be less related to risk, by transferring, packaging, and layering. Private credit hides risks through opaque valuations, liquidity mismatches, PIK structures, and similar methods, forming a structural foundation of industry fragility. And because software assets are being repriced due to AI disruption, they become the triggering factor that punctures the illusion of stability. In a loose-cycle environment, this risk-hiding mechanism appears to improve financing efficiency; but once a pressure cycle begins, its drawbacks become fully visible: risk exposure is delayed, and the transmission path is complex and opaque. Powell believes current risks are not yet a systemic crisis; his core focus is on whether they will infect the banking system. Bank of England Governor Bailey, meanwhile, warns against underestimating the risk-amplifying effect of “opacity plus interconnectedness.”

3 Impact: This is still a liquidity-and-trust crisis, not a full solvency crisis; be alert to what’s coming

Bank of England Governor Bailey recently warned that default events emerging in the private credit space should not be taken lightly as isolated cases. The industry’s high opacity may amplify the shock, and its risk characteristics are reminiscent of the 2008 global financial crisis.

This crisis is gradually spreading along the transmission chain from private credit funds to multiple stakeholders—banks, insurance companies, real-economy enterprises, and investors. What exactly is the impact?

First is the direct shock to the private credit industry itself: the private credit industry is facing a run on redemptions, tighter financing, and a sharp drop in stock prices. Redemption restriction mechanisms are essentially “blocking the exit.” They prevent a redemption run from spreading, but they also reveal the true fragility of liquidity. In the future, the difficulty of raising new funding will rise significantly, and industry growth will slow markedly. Banks have already begun tightening loan standards for private credit funds. JPMorgan has started restricting loans to some private credit funds, and other banks will likely follow. As funding sources for private credit narrow, financing costs will rise accordingly. Since the end of last year, the interest rates banks require for the main debt sources of these funds (including BDCs) have increased. Funds with excessive exposure to technology will continue to face pressure. Smaller managers may be eliminated or acquired, and faster industry consolidation alongside increased redemption restrictions will become this industry’s “new normal.” In the first quarter of this year, share prices of private credit industry giants such as Blackstone, Apollo, Blue Owl, and Ares fell by 22% to 38%.

Second is banks: as ties between the banking system and private credit grow tighter, banks may face “reverse erosion.” On the surface, banks may look like merely “providers of capital” or “leverage partners” for private credit. But once problems emerge in the underlying assets, risk can erode the banking system through multiple channels, such as credit tightening, collateral depreciation, and pressure on capital adequacy ratios. The IMF also warned in its Global Financial Stability Report that non-bank financial institution exposures of many banks have exceeded their Tier 1 capital. The fragility of non-bank financial institutions can be quickly transmitted to the core banking system, amplifying the shock and increasing the complexity of crisis management. The Nasdaq KBW bank index has fallen nearly 10% since early 2026, far exceeding the 2% decline of the S&P 500 over the same period.

Third is investors—especially insurance companies, for which private credit holdings account for 35%—facing dual pressure on their balance sheets. Over the past few years, marketing messages about “high yield, low volatility, and periodic exits” brought many high-net-worth individuals and institutional investors. In pursuit of higher returns, U.S. life insurance companies have continued to increase their allocation to private credit funds. According to an IMF report, private credit is about 35% of North American insurers’ investment portfolios. A large amount of insurance capital continues to flow into private credit funds through complex structures such as “Insurance Asset Management companies (IAM).” In a low interest-rate environment, this allocation appears reasonable: trade liquidity for yield and opacity for stability. But now insurers’ balance sheets face dual pressure: on one hand, write-downs on private credit assets directly reduce solvency, forcing insurers to absorb a major shock that can directly affect the interests of policyholders; on the other hand, large-scale redemption requests compel insurance institutions to sell assets at a discount or incur liquidity losses—this issue is particularly prominent in the North American market.

Finally, it impacts real-economy enterprises. Private credit is an important funding source for many small and mid-market enterprises. These firms often cannot easily enter public bond markets or obtain loans from large banks; private credit is nearly their lifeline. If the industry continues to contract, these businesses will face rising financing costs and difficulty refinancing, further intensifying the negative feedback loop of crisis transmission and weakening the U.S. economy’s growth momentum.

Overall, what we have right now is still only a liquidity-and-trust crisis, not a full solvency crisis. The 5% redemption cap effectively blocks a run from spreading across the entire financial system. Using default scenario stress testing as a framework, Goldman Sachs pointed out that even in an extreme scenario where the default rate rises to 10%, the drag on GDP would be only 0.2% to 0.5%. In the baseline scenario, the private credit default rate rises from about 1% in 2025 to 3% to 4%, corresponding to the low-end range of leverage loan default rates in historical credit cycles—overall risk remains within a controllable range.

But the warnings from Bailey and Buffett also cannot be ignored: when opacity and interconnectedness stack together, panic itself may become the crisis. Once panic sentiment self-reinforces and forms a stampede-style wave of redemptions, this “gray rhino” could break through the fence at any time, and the resulting cascading impact would far exceed current expectations.