Debt Peak Meets AI Transformation: Bain Capital Warns of Possible Wave of Defaults in Software Industry

Bain Capital recently warned that with the continuous expansion of the disruptive impact brought by artificial intelligence (AI) technology, coupled with the approaching peak of debt repayment, the software industry faces the risk of a double-digit increase in loan default rates. Angelo Rufino, head of Bain Capital's North America Special Situations Department and European Enterprise Special Situations Department, said in an interview, "Market pressures are about to show. This is just a typical credit cycle phenomenon - an industry is initially overhyped and then flooded with huge amounts of capital." Rufino predicted that the default rate in the software industry could soar to the high single digits to low double digits. In comparison, the maximum expected leveraged loan default rate in the United States this year is only 5%, consistent with 2025. Currently, Wall Street has frequently sounded the alarm, warning that AI as a new productivity tool may not only impact the software industry but also reshape the financial services and asset management sectors. Rufino's views are consistent with previous statements made by Bruce Richards, Chairman of Marathon Asset Management LP. Richards had earlier stated that the software industry default rate in the private credit sector could reach 15% - about three times his expected overall direct loan default rate and similar to the worst-case scenario given by UBS analysts. As an investment institution deeply involved in the software sector, Bain Capital holds assets such as Rocket Software Inc. Although the latter's debt has been under pressure recently, Bain revealed that its exposure to the software industry in the special situation business segment is less than 5%, without disclosing the total amount of software debt holdings. Rufino believes that although many software service companies have stable subscription revenues and offer practical products at low costs, the rise of AI will limit their ability to negotiate price increases. This will drag down enterprise valuation multiples and supported enterprise value, making refinancing debt more difficult. "As the credit cycle evolves comprehensively, the market will be forced to realign capital structures to match the profitability of these business models," Rufino said. "It can be certain that many companies will face refinancing difficulties." However, he also stated that given years of deleveraging and the steady growth of the US economy, the crisis in the software industry is unlikely to spread into a systemic credit market issue. "This storm is likely to be limited to a few specific industries, and I believe the overall credit market will not see a significant rise in default rates," Rufino said. But he also believes that considering the potential market risks, current credit spreads are still too narrow. High-yield bonds have a premium of about 300 basis points relative to US Treasury bonds, which "from a risk-return perspective, is not worthwhile."