Earlier Monday, I mentioned that US equities could prove more sensitive to any rate hikes Kevin Warsh is inclined (or allowed) to deliver due to the capital-intensive nature of the current cycle.

That simple observation was the centerpiece for the latest weekly from Goldman’s Ben Snider, who suggested that notwithstanding the bank’s house view for on-hold rates, a pivot to hikes could be more challenging for stocks than “usual,” where “usual” refers to S&P performance around the onset of the last seven hiking cycles.

So far, there’s little evidence to suggest capital markets are ready to revolt in the face of massive hyper-scaler debt issuance (some $225 billion of which since September was digested easily), nor that anyone’s inclined to balk at a dramatic increase in equity supply (from IPOs and follow-ons).

That said, rate hikes on top of those delivered across the Powell hiking cycle could push WACC for corporate America to the highest levels in three decades excluding the spike in and around Lehman, as illustrated on the left, below.

The figure on the right’s a rather stark reminder: Hyper-scaler net debt went from negative $200 billion pre-ChatGPT to in excess of positive $200 billion as of Q2 2026.

That’s not just one helluva turnabout, it’s indicative of a complete business model metamorphosis.

“The cost of capital is becoming increasingly important as the magnitude of hyperscaler capex reaches 100% of cash flows from operations this year, incentivizing increased debt issuance to fund spending,” Snider wrote, adding that “the recent importance of the AI investment boom in boosting corporate earnings and share prices should make stocks even more sensitive to changes in the cost of capital.”

As a quick reminder: This is all that matters. Almost. AI infrastructure stocks are seen accounting for at least half of S&P 500 profit growth this year, and the same stocks comprise more than 40% of overall market cap.

The figure on the right shows the impact of hyper-scaler capex on a common metric for overall corporate business outlays (the ratio of capex to sales).

This is a macro issue too: US GDP growth’s now heavily dependent on AI-related business spending, and to the extent that’s funded in part by debt, a higher cost of capital could, as Trump put it last month, “kill success.”

And yet, as I’ve been over again and again of late, capex booms tend to presage rate hikes given the read across from business investment bonanzas for the near- and medium-term neutral rate, and the extent to which failing to hike as neutral moves up is to countenance passive easing and thereby risk inflation.