Chris Ratcliffe/Bloomberg via Getty Images
Peter Oppenheimer, Goldman Sachs’ chief global equity strategist, is drawing a distinction that’s becoming central to the debate over whether tech stocks are in a bubble: it isn’t valuations that worry him, it’s earnings.
Back in August, Oppenheimer put it plainly: “there does not appear to be a valuation bubble, but there may be an earnings bubble.” By September he had hardened that view, tying the risk to a specific mechanism, AI infrastructure spending and government borrowing competing for the same pool of capital, a dynamic he argues is pushing up the global cost of capital more broadly.
The hard data behind his concern is stark. Capital spending among AA-rated tech issuers jumped 65% year-over-year in the second quarter. U.S. convertible bond issuance has reached $135 billion so far this year, with AI-linked borrowers accounting for 44% of that total. Goldman itself has responded by raising its full-year forecast for U.S. investment-grade issuance by $200 billion, to a record $2.3 trillion, with AI-related issuers now making up a full quarter of all new supply.
Torsten Slok, chief economist at Apollo Global Management, has echoed the warning from a different angle, pointing to how that debt has performed once issued: “most of the paper issued in 2026 trades wider today than where it priced,” he said, a sign that investors are already demanding more compensation for the risk tied to AI-fueled borrowing.
Oppenheimer stopped short of calling this a repeat of 2008, pointing to three factors he says separate the current moment from a classic credit bust: robust profits at major tech firms, strong balance sheets across the sector, and AI demand that is still outstripping available supply. The risk he’s flagging isn’t that the technology lacks real economic value, it’s that the earnings being priced in to justify current spending may prove harder to deliver than the market currently assumes.
