While Wall Street remains unsettled by the sharp decline in software stocks, UBS Group has issued a new warning: credit markets may be the underestimated "hidden powder keg" in the wave of AI disruption. As artificial intelligence technology evolves at a pace far exceeding expectations, software and data service companies burdened with heavy debt—particularly those held by private equity—are teetering on the brink of default.
In a research report released on Wednesday, February 12, Matthew Mish, Head of Credit Strategy at UBS, stated plainly that markets are repricing for a "rapid and aggressive disruption." He forecasts that by the end of next year, the leveraged loan and private credit sectors alone could see an additional $75 billion to $120 billion in defaults. This estimate is based on UBS's baseline scenario: leveraged loan default rates are expected to rise by 2.5 percentage points, affecting roughly $1.5 trillion in loans, while private credit default rates could increase by 4 percentage points, impacting about $2 trillion.
"Markets have been slow to react because they truly didn't anticipate this happening so quickly," Mish noted in an interview with CNBC. He pointed out that with companies like Anthropic and OpenAI releasing their latest models, market expectations for the AI disruption timeline have been drastically compressed. "People are having to completely rethink how they assess credit risk related to this disruption, because this is no longer a problem for 2027 or 2028."
From a "growth narrative" to a race against time, investor logic regarding AI has undergone a fundamental shift this month. The market no longer views the technology as a universal boon for all tech companies, but rather as a brutal shakeout where winners take all. Although software stocks bore the initial brunt of the sell-off, panic has quickly spread to seemingly unrelated sectors like finance, real estate, and trucking.
Mish emphasized that in the face of AI's impact, companies can be clearly divided into three tiers: - The first tier consists of creators of foundational large language models, such as Anthropic and OpenAI. These are currently startups but are highly likely to rapidly emerge as the next generation of major public companies. - The second tier includes investment-grade software firms like Salesforce and Adobe. They possess strong balance sheets and ample cash flow, giving them the capacity to swiftly deploy AI and fend off challengers. - The third tier comprises software and data service companies owned by private equity. These businesses generally carry high debt levels and rely heavily on traditional business models, leaving them most vulnerable to AI-driven disruption.
Beyond the baseline scenario, UBS also outlined a more severe "tail risk" picture. Under this scenario, default rates would double the baseline estimates, and financing channels for a large number of companies would be cut off. "The ripple effect would be a credit crunch in the loan market," Mish described. "You would see a broad repricing of leveraged credit, and the credit shock would impact the entire system." This situation would resemble the junk bond sell-off in energy companies a decade ago or the credit freeze following the dot-com bubble burst over twenty years ago.
UBS analysts noted that while risks are accumulating, the actual path of development still depends on several key variables: the pace of AI adoption by large enterprises, the rate of improvement in AI models themselves, and market refinancing needs. Currently, approximately 20% of leveraged loans and private credit face refinancing pressure by 2028, meaning risks will continue to build over the next two years. "We are not yet calling for the tail risk scenario, but we are moving in that direction," Mish admitted.
It is noteworthy that this warning focuses on leveraged loans and private credit, the riskiest segments of the corporate credit market. These instruments typically finance sub-investment-grade companies, which are often backed by private equity and carry high leverage. As AI tools begin to erode traditional Software-as-a-Service (SaaS) business models, the cash flows of these highly indebted companies are facing unprecedented pressure. Market concerns suggest that if these firms cannot pivot quickly enough amidst technological change, they could become the first casualties of this technological revolution, with the ultimate cost being borne by the credit markets holding this massive debt.