Banks Tighten Lending Standards for AI Debt Deals
Banks and credit investors are increasing scrutiny of AI-related debt following a Nasdaq 100 selloff and concerns over post-election data center oversight.
Banks and credit investors are increasing scrutiny of AI-related debt deals and showing reluctance to expand exposure to borrowers. This shift follows a Nasdaq 100 correction that erased 2.5 trillion dollars in market value and growing anxiety over potential data center oversight following the U.S. midterm elections.
OCBC Group and other lenders are now stress-testing cash flows and questioning exposure levels, even for deals guaranteed by major technology firms such as Google and Anthropic. The pullback primarily impacts riskier borrowers, including neocloud companies and standalone data center operators. For example, Nebius Group shelved a planned high-yield bond offering, though it entered a 3.5 billion dollar loan facility with Citigroup.
Despite the current caution, Goldman Sachs Research projects that global hyperscaler debt issuance will rise to 420 billion dollars by 2027. However, investors are becoming more selective, and spreads on hyperscaler debt have widened as the market adjusts to shifting sentiments regarding AI growth.