The Bridge Ep. 17: Is Private Credit's Golden Age Over?

Summary

David Manlowe says private credit is no longer being rewarded as a single, broad asset class the way it was for most of the last 18 years; instead, performance is separating by manager, with software exposure now one of the biggest differentiators. He argues AI will create both winners and losers in software and other industries, so lenders are building explicit AI frameworks into underwriting and focusing on how business models may look over the next three to four years. He highlights 2027 as a key refinancing test for loans written during the boom, especially as public software valuations have already compressed versus prior underwriting assumptions. On product structure, he defends the 5% redemption cap on semi-liquid evergreen funds as necessary to preserve the illiquidity premium, and says higher liquidity would mechanically lower returns. Despite headlines around redemptions, he says institutional allocators still want private credit, spreads have improved, and investor attention is shifting toward Europe and infrastructure credit as AI-related capex reshapes capital demand.

Notes

• Private credit is shifting from a broad outperformance regime to manager-level dispersion.
• Manlowe says the asset class was favored for 17 of the last 18 years.
• He describes the industry as being in a temporary “doghouse” after recent negative headlines.
• Manager selection now depends heavily on software exposure.
• Average private credit platforms reportedly have about 25% software exposure.
• Benefit Street Partners says it has less than 10% software exposure.
• That lower software concentration is framed as an advantage in a dispersing market.
• AI is expected to create both tailwinds and headwinds for company performance.
• Some midsize software businesses may be displaced by AI-enabled coding and automation.
• Software loans were commonly underwritten at 25% to 30% loan-to-value.
• That underwriting implied roughly 70 cents of cushion beneath the loan.
• He says AI will produce winners, losers, and some complete busts.
• Terminal values for software companies have already compressed materially.
• Public software comps are now described as trading around 7x to 8x versus about 15x underwriting assumptions.
• He thinks 2027 is when the new underwriting multiple will become visible.
• He expects older software loan assumptions to look too rich by next year relative to the new market standard.
• Borrowing costs for software names are already rising.
• AI-driven diligence is being folded into underwriting across industries, not just software.
• Five- to seven-year loans require a three- to four-year view on business-model change.
• Industrial services are highlighted as an AI winner.
• Data centers and infrastructure buildouts pull in water, power, and wiring vendors.
• Industrial services are portrayed as geographically dispersed and less disrupted by AI.
• Underwriting now places more weight on the management team.
• He frames memory chips as an unexpectedly hot commodity.
• Private credit flavors have multiplied across asset-backed, investment grade, infrastructure, and opportunistic strategies.
• Institutional investors are expected to keep allocating to private credit.
• Institutions liked the first flavor because it outperformed public credit over long periods.
• He says there is no diminution in institutional demand.
• Redemption pressure is concentrated in new wealth-platform products, not institutions.
• He says the product was mislabeled as semi-liquid and should have been called illiquid.
• The 5% redemption cap is presented as a protective feature for all investors.
• He argues illiquidity is what earns the premium in private credit.
• Adding 10% to 20% liquidity would reduce yield and spread.
• He says loan spreads widened from four-handle levels to five- and six-handle levels.
• He expects wealth-channel repricing pressure to persist until around 2027.
• Performance is the main cure for redemption pressure.
• The broadly syndicated loan market is expected to stay wide open.
• He sees strong capital demand in the core middle market from AI and hyperscaler projects.
• Large AI projects could siphon capital away from competing lenders and lower competition in middle-market lending.
• Diversification away from crowded credit exposures may benefit private credit generally.
• He says Europe is attracting more attention because U.S. allocations are already heavy.
• Infrastructure credit is the current “hot dot” for investors.
• He sees a long 5-, 10-, 15-year infrastructure super-cycle.
• AI is also creating tailwinds for infrastructure investment.
• He rejects a 16% to 17% default-rate narrative as unrealistic.
• His internal read is that default rates are around the three-handle, possibly 4% to 5%.
• Geopolitical risk is one of his main macro worries.
• He is also concerned the Fed could hike rates again and hurt the 3-4 year horizon.