Low Loan-to-Values Shield Private Credit From the AI Selloff

What's New

AI disruption threatens the exit multiples sponsors paid for software, while low loan-to-values keep lenders' recovery largely intact. David Hirschmann, Co-Head of Permira Credit, argues this in a podcast interview on Fund Shack. Sponsors often pay 20 times EBITDA or more for software businesses. Lenders typically fund 5 to 6 times. European direct lending loan-to-values run around 30 to 40%, so lenders are repaid in full if a third to 40% of the entry value survives. LPs should judge software credit on cash flow durability over a 4 to 5 year loan life.

Why It Matters

The argument splits a single AI trade into two risks with different owners. Sponsors carry multiple compression. Lenders carry refinancing risk on businesses that still generate cash. UBS sits across the table with a forecast of 15% private credit defaults, which Hirschmann calls a pessimistic scenario that would hit all credit, private and public. Permira's opportunistic credit strategy buys discounted loans in the liquid market, so a thesis that the selloff is indiscriminate supports its own trade.

What To Watch

• Secondary loan prices for vertical, data-rich software businesses. Hirschmann says the liquid market is selling these alongside easily replicated horizontal tools. A widening price gap between the two would show the market sorting credits the way he does.

• Sponsor equity injections in stressed deals. His recovery model relies on owners funding new money to keep control. More handovers of the keys after one or two failed injections would test it.

• Redemption requests at semi-liquid private credit funds. He ties the current fear to these vehicles and AI headlines, and Permira does not depend on monthly inflows.

Memorable Quotes

• "If we manage to recover a third or 40% of the initial enterprise value, we're money good." Low entry leverage turns distress into a manageable workout for lenders.

• "All software businesses are doomed and are going to disappear. That's also wrong." Hirschmann treats today's panic as the mirror image of the old view that software was the safest credit.

• "We're talking about a small number of situations, but in some cases we've recovered more than the nominal." His above-100% average recovery rests on a thin sample of defaults.

• "Whether the exit multiple is 15, 17 or 20 times will not have an impact on my return or my default or loss rate as a lender." Equity and credit investors in the same company are exposed to different outcomes.

The Wrap

Hirschmann's view works while AI erodes software revenue gradually and margins stay high enough to refinance. It breaks if recurring revenue collapses within a single loan term. It also breaks if recoveries fall below the 30 to 40% cushion that equity provides at entry. A larger default sample would test Permira's recovery record. Software loans written at peak valuations come up for refinancing over the next 4 to 5 years, and that wall will settle it.