The Rising Significance of Capital Expenditure in the AI Era

In the current earnings season, the traditional focus on earnings per share is being challenged by a more critical metric: capital expenditure (CapEx). Andrew Sheets, Global Head of Fixed Income Research at Morgan Stanley, highlights that the massive investment in AI infrastructure—the "chips, power, cooling, and connections" necessary to run AI models—has become the defining narrative for the world's largest companies.

The Unprecedented Scale of Investment

The sheer volume of capital being deployed is staggering. Morgan Stanley estimates that the largest US hyperscalers—including Alphabet, Amazon, Microsoft, and Meta—will spend over $600 billion this year on capital and equipment. To contextualize this, Sheets notes that this handful of tech giants is set to spend as much as "every non-technology company in the S&P 500" combined.

This trend is not merely a one-time surge but an accelerating phenomenon. Forecasts have been repeatedly revised upward; just a year ago, the 2026 spending forecast was roughly half of current expectations. Looking further ahead, Morgan Stanley estimates that annual capital spending by US hyperscalers could reach $1 trillion by 2028, suggesting that the "spending story still lies ahead."

Market Winners and the Credit Dilemma

The implications of this investment cycle are twofold, affecting both equity and credit markets:

1. Equity Upside: Much of this spending flows directly into the revenue streams of AI suppliers. The market has responded accordingly, with "U.S. semiconductor stocks" rising over 30% in a single month, reflecting how one company’s massive CapEx becomes another company's revenue.

2. Credit Pressure: Despite the immense financial resources of these tech giants, the scale of investment necessitates significant borrowing. This has led to "record bond issuance" this year, with the first quarter marking the busiest period on record for U.S. investment-grade bond issuance.

This creates a precarious dilemma for investors. If tech companies remain confident and continue to raise their CapEx guidance, it bolsters the equity market but forces the corporate bond market to absorb even higher levels of debt, which Sheets identifies as "credit negative." Conversely, if companies guide spending down, it could "undermine a key pillar of recent market strength," potentially dampening broad risk appetite. Given this, Sheets suggests that the risk-reward profile is currently more attractive in other areas of fixed income, such as "mortgage-backed securities."

Macroeconomic and Monetary Policy Implications

Finally, the long-term success of this capital deployment may influence the Federal Reserve. Drawing on the perspective of Kevin Warsh, Sheets notes that high levels of investment are theoretically capable of boosting productivity, which could "lower inflation" and justify a lower interest rate environment. Ultimately, the confidence these hyperscalers exhibit regarding their AI investments—and the tangible results that spending delivers—will play a pivotal role in the broader "monetary policy story."

As the earnings season progresses, the market remains caught between the transformative potential of AI infrastructure and the tightening constraints of the corporate credit market.