On November 18, Robert Duffern, a senior executive at Polus Asset Management, noted that the debt boom in the AI sector reflects a marked deterioration in credit market lending standards. Compared with 25 years ago, many borrowing companies today carry higher leverage and lower-quality earnings and cash flow. Some companies rely primarily on intangible assets to secure financing, while venture capital-backed firms have accumulated substantial debt with thin equity buffers. He warned that these issues will exact a cost in the years ahead. Data show that since 2025, global companies have borrowed more than $820 billion for AI buildout, yet the returns, realization timelines, costs, and technological applications of these investments remain uncertain. Duffern believes the AI financing boom is crowding out capital from other industries, may heighten companies' future refinancing difficulties, and could drive more debt restructurings. JPMorgan expects the default rate on leveraged loans to rise to 4.5% next year—more than double the current level—and the default rate on high-yield bonds to climb to 2.75%.
