Goldman Sachs research indicates that artificial intelligence investment is fundamentally altering how U.S. corporations raise capital, while noting that evidence of AI spending crowding out other business investment remains limited for now.

The bank projects U.S. AI investment will reach nearly $600 billion in 2026, equivalent to approximately 2% of GDP. AI investment has accounted for over 10% of business fixed investment in recent quarters—a concentration that recalls prior technology capex cycles such as the late-1990s telecom buildout, though Goldman notes the current surge is driven by a narrower group of companies with substantially stronger balance sheets.

Record Equity Issuance Driven by AI Sector

U.S. corporations raised $252 billion through IPOs, follow-on offerings, convertible securities, and SPACs in the second quarter of 2026, surpassing the previous record of $234 billion set in Q1 2021. That prior peak coincided with the post-pandemic SPAC and retail-trading boom, meaning the current record is being set under markedly different market conditions—driven primarily by institutional capital flowing into AI infrastructure rather than speculative retail demand.

AI-related companies were responsible for roughly 40% of all U.S. follow-on equity volume this year. Technology, media, and telecom firms represented 28% of follow-on volume, more than double their average share over the previous five years.

Goldman Sachs identified Amazon, Alphabet, Meta Platforms, Microsoft, and Oracle as the primary hyperscalers driving the spending surge. Consensus estimates place their combined capital expenditure above $1 trillion annually over the next several years.

The dependence on a handful of companies means the broader financing outlook hinges on the capital allocation decisions of five firms. Capital spending by these companies is projected to exceed their operating cash flow by approximately $150 billion in 2027. Should spending reach $1.4 trillion as some investors anticipate, the funding shortfall could surpass $300 billion.

Goldman Sachs strategist Ben Snider characterized the increase in equity issuance as a return to normal levels rather than a sign of financial strain. Total issuance represents only about 1% of the Russell 3000's market capitalization, roughly consistent with the 2015–2019 annual average.

Debt Markets and Buybacks Expected to Absorb the Supply

Debt is expected to bear the majority of the financing burden. Goldman Sachs credit strategists forecast that hyperscalers will fund 35% of their 2027 capital spending through debt, corresponding to approximately $400 billion in global debt issuance next year.

Regarding potential crowding-out effects, Goldman found only limited evidence that AI spending is displacing other business investment. While AI-related financing has grown to nearly one-quarter of investment-grade issuance, non-AI credit spreads remain near historical lows—an indicator that traditional sectors continue to access capital on favorable terms despite AI's growing share of bond market activity.

Follow-on offerings have been priced at an average discount of approximately 7% to pre-announcement levels. Post-offering stock performance has remained close to historical norms, indicating that investors are absorbing the new supply without significant disruption.

Share repurchases are projected to more than offset the increase in new share issuance. Goldman estimates U.S. companies will buy back $1.4 trillion in shares this year. S&P 500 buyback growth was running at 11% year over year in Q2, and total buyback authorizations had reached a record $989 billion as of the report date.

Goldman expects $700 billion in total U.S. corporate equity issuance for 2026, a dollar record. The bank summarized the dynamic succinctly: equity issuance is a "headwind but not a gale."