The private credit expansion of the last decade moved through several phases—leveraged buyout debt, healthcare, real assets—but the 2022–2024 vintage was notable for its concentration in mid-market enterprise software. That concentration is now the source of the outflow pressure working through the largest perpetual credit vehicles, and the disclosure structure that defined the asset class is making the problem harder to resolve.
The CEPR Analysis That Named the Architecture
Center for Economic and Policy Research co-director Eileen Appelbaum laid out the structural chain in April 2026. PE firms had spent the prior seven years acquiring life-insurance and annuity businesses and redirecting policyholder reserves into proprietary private credit funds. The funds operated with thin disclosure requirements and infrequent marks. They deployed that capital into PE-owned portfolio companies, with significant concentration in mid-market application software businesses during the period of peak enterprise software valuations.
The structural concern Appelbaum identified was not simply that the exposure exists. It was that the chain of capital—from policyholder, through insurance intermediary, through PE-controlled credit fund, to software borrower—insulates the ultimate risk bearer from information about what that risk actually is. LPs in these funds cannot answer, from disclosed data, how much of their allocation is in AI-substitutable software categories.
What the Funds Disclose and What They Do Not
Standard private credit fund disclosures include sector allocations, vintage distributions, and aggregate credit statistics. Software appears as a category total—typically 20% to 35% of AUM at the funds most active in the 2022–2024 period. Within that category, no major fund currently discloses the split between infrastructure software (AI-resistant), vertical SaaS (variable), and horizontal application software (highest AI-displacement risk). The AI-displacement-risk metric does not exist as a disclosed category.
The absence of that metric is what is driving exits. When LPs cannot quantify their exposure to a specific risk category, and when that risk category is material and uncertain in its timeline, the default response for institutions with risk mandates is to reduce the allocation. Redemption volumes climbed through the fourth quarter of 2025 and continued rising in the first quarter of 2026.
Three Gates in Six Weeks
Two of the largest perpetual private credit vehicles disclosed quarterly outflow caps in March 2026. A third moved in early April. All three announcements came without disclosure of material credit losses—the gates are responding to redemption queue volumes, not to confirmed borrower distress. Secondary market buyers have independently moved the discount on fund interests to reflect the possibility of future mark-downs, creating a spread between stated NAV and secondary clearing prices.
Portfolio Risk Is Not Evenly Distributed
The AI-displacement risk concentrates in specific portfolio characteristics, not across the asset class uniformly. Funds that lent most aggressively to horizontal application software—productivity, CRM, project management, document automation—between 2022 and 2024 carry the highest density of potentially vulnerable borrowers. Funds that built positions in infrastructure software, mission-critical vertical SaaS, or asset-backed structures are in a substantively different position.
The manager defense—tighter covenants, private workouts, no forced-sale mechanics—is a genuine structural advantage. It describes the workout process, not the credit outcome. The credit outcome depends on how many software borrowers experience AI-driven revenue declines severe enough to trigger covenant violations, and over what time frame those violations cluster. That is not a question any fund has yet answered, because the revenue data needed to answer it is still being written in 2026 enterprise software markets.
The signal that will clarify the picture most quickly: NAV prints from the largest perpetual vehicles over the next two quarters, and LP letters that begin carrying explicit AI-displacement-risk disclosures. The latter will arrive only after LPs demand it loudly enough to change what managers put in writing. Based on current redemption levels, that demand appears to be building.
Source: Private Credit Fund Redemptions Climb Sharply, Some Caps Now in Place
