AI stock dip masks deeper bond market pushback on tech debt
A recent sell-off in AI-linked equities has been driven by leveraged liquidations, but the more consequential shift for corporate finance is cooling demand in the bond market.
US technology stocks fell this week, leaving the S&P 500 down 1.6% and the Nasdaq 2.9% lower, after a new Chinese AI model cast doubt on the massive capital expenditure plans of hyperscalers. While equity strategists view the stock pullback as a healthy consolidation, the selloff points to a more structural challenge: a growing reluctance among bond investors to finance the AI boom at current prices.
Fundstrat Global Advisors cofounder Tom Lee noted that margin debt has surged 54% annually, marking the sixth largest increase in six decades. He attributed the swift decline in chip stocks to "push-button liquidity for everybody," where leveraged funds and zero-day options force rapid liquidations. Lee argued the companies driving the AI trade remain central to US strategic interests and still have years of runway. “I would still stick with those,” Lee said. “I think those names are going to bounce later this year. So I don’t think that the trade is over.”
The dangers of that leverage are already visible in Asia. South Korea's Kospi index has plummeted 27% from its record high last month, entering bear market territory. Lee estimated that 1.2 million Korean brokerage accounts, roughly 10% of the total, faced margin calls. The unwind was triggered in part by SK Hynix announcing plans to slow its AI memory business, which sparked the Kospi's fifth-worst daily plunge and dragged global indexes lower.
Corporate Debt Costs Rise
Compared to the Korean crash, the S&P 500's 2% dip and the Nasdaq's 6% slide look like minor corrections. However, the fixed-income market is sending a more stringent signal. Apollo Global chief economist Torsten Slok noted that the hyperscaler cover ratio has collapsed from nearly 5x to below 2x, “suggesting investors may need wider spreads to absorb additional hyperscaler supply.”
While stock investors chase AI trends, bond buyers are demanding higher compensation for the risk. Tech companies are spending hundreds of billions of dollars annually on data centers, saturating the world's largest dollar bond market. Issuers are now being forced into other currencies to find buyers, a dynamic that will inevitably raise their borrowing costs.
This financing squeeze comes at a difficult time. AI-related corporate debt must now compete for capital against a massive wave of US Treasury issuance, as the federal deficit is on track to hit $2 trillion this fiscal year. “The current hyperscaler widening is a byproduct of the high-grade investor community trying to rationally price in an accelerating pace of issuance,” JPMorgan strategists wrote.
