A structure that regulatory frameworks have been slow to reach is now at the centre of the private credit outflow story. Private equity firms control significant insurance and annuity operations, whose policyholder reserves underpin private credit funds that have lent heavily to mid-market software companies. That chain—insurance capital through PE intermediary to leveraged software debt—is generating the LP exit pressure that resulted in three fund gates between March and April 2026.
CEPR Identifies the Architecture
Eileen Appelbaum, co-director of the Center for Economic and Policy Research, published a structural analysis in April 2026 documenting how PE firms built this chain over the preceding seven years. The pattern: PE firms acquire life-insurance or annuity businesses, gain control of policyholder reserves, and redirect those reserves into proprietary private credit funds that operate with thin disclosure and no mark-to-market obligation. The credit funds then deploy the reserves—through layers of intermediation—into PE-owned portfolio companies, with significant concentration in mid-market software borrowers during 2022–2024.
The regulatory concern embedded in the CEPR analysis is that the investment risk borne by policyholders is significantly different from what traditional insurance regulation anticipated. UK and US insurance regulatory frameworks were designed around investment-grade bonds, public equities, and conventional diversification. Private credit concentrated in leveraged software debt—managed by the same PE entity that controls the insurer—does not fit those frameworks easily. The disclosure requirements that apply to the insurance entity do not extend in useful form to the credit fund’s underlying loan book.
How AI Displacement Exposed the Gap
Software was a consensus-safe lending category in 2022. The AI disruption risk that materialised in 2025 is categorically different from the risks that credit underwriting typically models: it is a structural question about whether specific revenue models remain viable as a new technology matures, rather than a cyclical question about whether borrowers can service debt through an economic downturn. Credit models built for cyclical risk do not answer structural technology displacement questions.
The disclosure gap compounds the problem. Private credit fund letters disclose aggregate software exposure. The sub-category distribution—between infrastructure software, vertical SaaS, and horizontal application software facing direct AI substitution—is not disclosed. LPs and, by extension, the insurance regulators overseeing the PE-controlled insurers, cannot measure the AI-displacement risk sitting in the portfolios that their capital backs.
Three Gates Without Losses
Two perpetual private credit vehicles imposed quarterly outflow caps in March 2026. A third followed in April. None disclosed material credit losses. The gate mechanism—standard in private credit fund documentation—allows managers to limit quarterly redemptions to a percentage of fund NAV, typically 5% to 10%. The mechanism was designed for normal exit management. At three of the largest vehicles simultaneously, in a six-week window, it signals an exit pressure that exceeded what the normal liquidity architecture was designed to handle.
The Market Pricing That Has Preceded Regulatory Action
Secondary buyers of fund interests have established discounts above stated NAVs. The spread widens with each gate announcement. The market is performing price discovery that regulatory oversight has not demanded from the fund disclosure architecture. That is a familiar pattern: private market stress surfaces through secondary transaction pricing before it appears in regulatory data or formal credit metrics.
The UK regulatory position on PE-controlled insurance capital is under active review following the CEPR analysis and related work from other policy research institutions. Whether that review produces disclosure requirements specific enough to address AI-displacement risk in software loan portfolios—at the sub-category level that would allow meaningful risk assessment—is the regulatory question the current cycle has made unavoidable.
NAV prints from the largest perpetual vehicles over the next two quarters and any emergence of AI-risk metrics in LP letters will provide the first credit-market evidence. The regulatory evidence will take longer to accumulate, as it always does.
Source: Private Credit Fund Redemptions Climb Sharply, Some Caps Now in Place