The Earnings Bubble Beckons: A Cautionary Tale of Capital’s Collision Course

Goldman Sachs’ chief global equity strategist, Peter Oppenheimer, has added hard numbers to his earnings-bubble warning, tying the risk of an AI-driven earnings bubble to a specific mechanism: the collision course between government borrowing and private capital spending. This collision is not just theoretical; it’s playing out in real-time.

Capital expenditures among AA-rated technology issuers surged 65% year-over-year in the second quarter, while record-breaking $2.3 trillion in U.S. investment-grade issuance this year has AI-related issuers accounting for a quarter of that supply. Oppenheimer’s report is part of a larger trend, as several senior Wall Street voices have converged on this framing.

Torsten Slok, chief economist at Apollo Global Management, argues that what was once a “savings glut” has turned into a “savings shortage,” where projects are abundant but capital is scarce. This competition for capital is driving up the global cost of capital, as evidenced by widening spreads on hyperscalers’ longest-dated bonds and the fact that most paper issued in 2026 trades wider today than when it priced.

Investors are still buying, but they’re charging more – a phenomenon particularly visible in secondary bond markets. This trend is not just a repeat of prior instances where earnings blew up first; today’s technology sector looks different from its predecessors. Profits remain robust, balance sheets are strong overall, and interest coverage ratios for the aggregate S&P 500 are healthy.

However, what happens when the music stops? When the capital tide recedes, leaving behind a landscape of struggling tech companies with bloated balance sheets and declining earnings growth? The consequences could be severe, particularly if investors have been pricing in unrealistic expectations of future growth.

The 2008 financial crisis offers a cautionary tale for today’s tech sector. Banks briefly became the largest sector in the S&P 500 in the run-up to the crisis, but their earnings were inflated by rapidly rising leverage financing an asset that experienced a genuine valuation bubble: U.S. real estate. When housing collapsed and pushed the economy into recession, bank earnings collapsed with it.

Oppenheimer notes that technology profits remain robust and balance sheets are strong overall – a contrast to the credit-fueled fragility that eventually undid bank earnings. But what happens when interest rates rise and capital becomes scarcer? Will tech companies be able to sustain their earnings growth, or will they become the next victims of an earnings bubble?

The surge in AI-related issuers is driving up the global cost of capital, as governments compete for capital by borrowing more for infrastructure, energy security, and defense. These companies are raising debt and equity to fund their data centers, but they’re also contributing to the broader trend of rising capital expenditures among AA-rated technology issuers.

The bond market is sending clear warning signs about the earnings bubble: spreads on hyperscalers’ longest-dated bonds have widened, and most paper issued in 2026 trades wider today than when it priced. Investors are still buying, but they’re charging more – a phenomenon particularly visible in secondary bond markets.

Long-term rates have moved more than short-term ones, with data centers, power generation, transmission, and government deficits all being long-duration claims on savings. This means that the competition for capital concentrates at the long end of the curve. As Slok notes, this is not just a sign of rising interest rates; there’s something more fundamental at play.

The earnings bubble beckons, and tech investors would do well to heed the warning signs. Oppenheimer’s report has sounded the alarm, but it’s up to investors, policymakers, and corporate leaders to take action. The clock is ticking – and the consequences of ignoring this trend could be catastrophic.