About $860 billion in net asset value is stuck in U.S. private equity funds that are more than seven years old, according to PitchBook data reported by Fortune. The figure comes from more than 4,500 U.S. PE-backed companies—33.8% of the 13,509 tracked—that have been held for five years or more.

Kyle Walters, PitchBook’s private equity analyst, defines the problem narrowly: capital that should have been returned to investors is instead locked in companies held for seven to ten years with no clear path to a successful exit. These are not merely old assets. They are assets bought when debt was cheap and valuations were high, and the exit math no longer works.

The origin was zero-interest-rate policy. Cheap debt fueled a buyout boom, particularly in 2020 and 2021. Walters describes companies bought at 12x that are now worth 10x. When rates rose to 40-year highs in 2023, the financial-engineering leg of the value-creation plan stopped doing the work. The remaining path is operational improvement, and it arrived at exactly the point when the economy made that hardest.

For a limited partner, the central question is not just age. It is what the sponsor was actually buying and which value-creation path was supposed to close the gap. A 2019-2022 vintage deal underwritten to multiple expansion or a cheap refinancing is now fighting the clock. A deal underwritten to EBITDA growth has a testable metric: did operating earnings grow enough to absorb a step-down from a 12x entry multiple to a 10x exit multiple? PitchBook’s aggregate figures do not name firms, but they imply a large number of GPs are holding assets where current market multiples may leave little or no equity cushion. That is an inference from Walters’s 12x-to-10x example, not a disclosed portfolio figure.

Walters does not see a systemic failure yet. His point is that GPs have timing discretion. They can extend hold periods, amend debt, or sell a zombie to a platform that wants it as an add-on. A structural crisis would require a second layer of stress — a refinancing wall, a covenant breach wave, or an LP liquidity shock — on top of the existing zombie inventory. The zombie inventory can persist for years as long as lenders and LPs are willing to accept delayed exits.

Walters expects the strong to acquire the weak; some zombie companies will become add-on acquisitions and achieve a delayed final exit. Fortune’s Allie Garfinkle points out the more severe path: some will go bankrupt or wind down. Walters says both outcomes are inevitable. "These companies can’t sit in the portfolio forever," he said. "They have to decay in one way or another."

The next test for an LP is not whether the $860 billion aggregate moves. It is whether the GP discloses how many portfolio companies have been held between seven and ten years, what multiple was paid, how much net debt sits on each asset, and whether EBITDA growth has been enough to absorb the decline in the entry-to-exit multiple. If those numbers are not available, the zombie may already be in the LP’s own portfolio.