Goldman Sachs Forecasts $600 Billion in US Stock Sales in 2027 After Record 2026

Corporate America is selling stock at a pace the US market has never seen, and Goldman Sachs expects it to continue.

In a report dated October 9, 2026, Goldman chief US equity strategist Ben Snider forecast $600 billion in US equity issuance for 2027. That would follow what the bank now expects to be an all-time record this year.

The numbers

Goldman splits the 2027 forecast into $175 billion from initial public offerings and $425 billion from follow-on offerings, convertible securities and SPACs.

Companies have already raised roughly $431 billion through equity offerings so far in 2026, up 98% from the previous year. Goldman expects the full year to reach $675 billion. The old record was $540 billion in 2021.

AI is the common thread. AI-related follow-on offerings have raised about $65 billion this year, around 45% of all US follow-on issuance, according to Goldman.

Spending that outruns cash flow

Goldman expects Amazon, Alphabet, Meta, Microsoft and Oracle to spend $1.2 trillion on capital expenditures in 2027. Operating cash flow for the same group is projected at $1.1 trillion.

A company that spends more than it earns has to borrow or sell shares to cover the difference.

A second supply wave is also coming. About $1.7 trillion in shares could become tradable in 2027 as lock-up restrictions expire. Goldman says that could produce the largest increase in net publicly available US equity supply since 2000.

The bank's view is that the market can handle it. Snider's report predicts that heavier corporate buybacks and sustained investor demand will absorb the volume and keep the bull market going, even if valuations come under some pressure.

How big the AI buildout is

New estimates from the Brookings Institution put the scale in historical context. Economist Stijn van Nieuwerburgh projects total investment in data centers and related AI infrastructure of $10.3 trillion between 2025 and 2032, averaging 3.6% of GDP a year. Goldman, using a different method, puts US AI investment at 1.9% of GDP in 2026 alone.

For comparison, van Nieuwerburgh's figures show railroads averaging 2.24% of GDP from 1870 to 1890 and the interstate highway system 1.13% from 1956 to 1973. His AI estimate tops both.

The spending already shows in Commerce Department data. Seasonally adjusted private data center construction reached $37 billion through July, about $9 billion more than the same stretch last year. Private construction spending on everything else ran roughly $46 billion below year-earlier levels over those seven months.

FactSet's analyst consensus has Google, Amazon, Meta, Microsoft and Oracle spending a combined $4.2 trillion over the four years ending in 2029.

The risks

Van Nieuwerburgh says more of this money is coming from debt. Tech companies often borrow through off-balance-sheet entities from banks and private-credit firms, he said, arrangements that carry little public reporting and make the scale of the financial risk hard to gauge.

There are local costs as well. The Federal Reserve Bank of Richmond has reported that data center construction is straining labor availability in its region. In Mississippi, a proposed aluminum smelter that would have brought an estimated 1,000 permanent jobs went to Oklahoma last year. A person familiar with the operators' decision-making said a data center announced near one of the sites under consideration, near Vicksburg, had tied up the electricity the smelter needed. Site-selection consultant Didi Caldwell, who works with heavy industry, said: "It's crowding out manufacturing."

Goldman's own research acknowledges the unease on Wall Street. Its September 23 note said recent earnings strength has raised concerns among some investors that stocks are in an "earnings bubble."

Earnings are doing the lifting

That same note credits AI investment with driving nearly half of S&P 500 earnings-per-share growth this year. S&P 500 EPS grew 51% in the second quarter from a year earlier and 26% over the past four quarters, according to Goldman Sachs Research.

Goldman expects that contribution to fade. It projects AI investment shifting from an earnings tailwind this year to a marginal drag in 2028, making actual productivity gains from AI more important for profits.

The bank forecasts S&P 500 EPS of $415 in 2027 and $460 in 2028. It targets the index at 8,700 over the next 12 months, up from 7,764 on September 21, driven by profit growth rather than higher valuations.

What to watch

The forecast rests on two assumptions: that buybacks and investor demand keep pace with $600 billion in new stock plus the $1.7 trillion lock-up expirations, and that AI productivity gains arrive as the investment tailwind fades. Goldman expects both. Whether they hold will be tested through 2027 as the lock-ups expire.