Private credit marks down software holdings as $1.6 trillion asset class reprices quietly

Private credit funds are marking down software company holdings across their portfolios, a scattered but unmistakable signal that the $1.6 trillion asset class is repricing assets beneath the surface while continuing to market stable returns to limited partners. The write-downs, disclosed piecemeal in fund reports over the past quarter, cluster in software businesses financed during the 2021-2023 credit boom—companies that carried aggressive revenue multiples and covenant-light structures when capital was abundant.

The pattern matters because software has been private credit's comfort zone for a decade. These were the borrowers deemed "asset-light but cash-rich," the recurring-revenue models that justified 12x-15x EBITDA purchase multiples and debt packages at 6x-7x leverage. Now several of those names are being revalued downward, not because of default but because the arithmetic of exit assumptions no longer holds. Direct lending issuance in the U.S. has slowed measurably in recent months, and the bid-ask spread between what sponsors will pay and what lenders will accept has widened past the point where deal flow can paper over aging positions.

What separates this from routine portfolio management is timing and breadth. These markdowns are appearing while the funds in question still report net asset values within 2-3% of par and while industry fundraising, though below its 2022 peak, continues at a $150 billion annual pace. The implication is that valuation discipline is being imposed selectively—on positions where exit paths have narrowed or where borrowers are burning cash faster than underwritten—but not yet systematically across portfolios. That selectivity creates a marking lag, and the lag creates opportunity cost for allocators trying to model true exposure. The software write-downs are canaries, not because software is uniquely fragile, but because it was uniquely overweight in private credit portfolios during the vintage years now coming under scrutiny.

The second-order effect is structural. Private credit sold itself to allocators as a yield pickup over liquid credit with downside protection via seniority and covenants. The software markdowns suggest that seniority is being tested in slow motion—not through headline defaults, but through quiet re-trades where lenders accept lower recoveries to avoid crystallizing losses or triggering fund-level writedowns that would spook LPs. This is the behavior of a market that has priced in permanence of capital but not permanence of assumptions. Ares launching a new Asia fund while U.S. issuance slows is consistent with this: the largest managers are rotating toward geographies where the valuation reset has not yet occurred, preserving the growth narrative while the home market digests.

Allocators should watch for two developments over the next six to nine months. First, whether Q3 and Q4 fund reports show markdowns broadening beyond software into other asset-light verticals—healthcare services, business services, anything financed on EBITDA multiples above 10x in 2021-2022 vintages. Second, whether the slowdown in direct lending issuance reverses or whether it continues, because sustained slowdown forces funds to hold positions longer and marks the end of the "refinance your way out" exit strategy that has kept reported NAVs stable. The funds with the most software exposure from 2021-2023 are the ones now managing the tension between what they told LPs the portfolio was worth and what the market will pay.

The valuation adjustment is not a crisis. It is the asset class learning to mark to reality instead of model.