Big Tech's AI binge ends the buyback era: judge growth, not payouts

What do you call a company that used to hand back about a quarter of its operating cash and now hands back less than a tenth? Not a bad company, necessarily. A company that has changed its mind about who gets paid first.

That is the story behind the latest data. According to a Yahoo Finance analysis, the five big spenders returned just 9.3% of operating cash flow as dividends and buybacks, down from 27.4%. That is the smallest share in at least 15 years. Separately, filing-level data compiled by Axis Intelligence says Alphabet and Meta bought back $0 of stock in the first half of 2026, to fund AI capex.

The cash went into AI spending. Platformonomics tallies that Amazon, Google, Microsoft and Meta spent more than $210 billion in the third quarter, up 84% from a year earlier. Alphabet's quarter shows what that looks like in practice: CNBC reports negative free cash flow of $9.07 billion, with revenue and earnings short of Wall Street expectations.

The buyback was never the product

For a decade, a big part of the Big Tech pitch was a quiet one: these companies print so much cash that they can fund the future and shrink their share count at the same time. Buybacks became a comfort blanket. In our view, even when growth cooled, the stock had a floor of corporate demand under it.

We think that floor is now thinner, and that is not a scandal. It is arithmetic. Epoch AI's analysis of SEC filings, as summarized by Let's Data Science, finds hyperscaler AI capex growing about 70% a year against roughly 23% growth in cash flow. Two lines like that cross. The same summary says free cash flow reaches roughly zero around the third quarter of 2026, and that Oracle is already negative. A company cannot return cash it does not have, and in our opinion it should not borrow to buy back stock while racing to build capacity.

So the real question is not "why did they stop?" It is "why did anyone expect them to keep going?" Shareholder yield looks like a byproduct of a capital-light era. This is a capital-heavy one.

Growth has to pay the bill now

If payouts are off the table, the stocks live or die on whether the spending earns something. The evidence so far is mixed, which is exactly the point. One third-party summary of Alphabet's quarter, from Nexuswild, has revenue of $127.6 billion, up 25%, and Google Cloud revenue more than doubling to $30.7 billion. It also puts capex at $51.4 billion and says the shares fell 10%, with capex and negative free cash flow overshadowing the cloud growth. Strong growth, and the shares fell anyway, in a quarter when revenue and earnings also missed expectations.

That reaction tells you how the market seems to have rewired. Headlines from the last few days, such as Moneycontrol's on AI spending concerns deepening a Wall Street selloff, show investors asking for proof, not promises. We would say that is healthy. Paying a premium multiple for a company that is simultaneously spending more than it generates and no longer returning cash demands a different kind of homework.

Own these stocks for growth you can see, not for payouts

Our view, and it is an opinion rather than advice: stop reading payout ratios as a quality signal for these names. A low buyback figure today may be a sign of ambition, not weakness. A high one, at a company with no visible AI return, might be the more worrying sign.

Three questions are worth more than the yield. Is revenue growth accelerating in the businesses the capex is meant to feed, such as cloud? Is free cash flow bottoming or still falling? And is management disciplined, with guidance it keeps, or does the spending number keep ratcheting up?

There is also an uncomfortable counterpoint. Some of these companies may be building capacity ahead of demand that arrives late, or never at the expected scale. Nobody can tell us yet which. That is why the risk sits with shareholders, who are effectively being asked to fund the bet through foregone returns.

The verdict: the buyback era for Big Tech is probably not dead forever, but it looks to be on pause, and that pause is the price of the AI strategy. If you own these stocks, own them for growth you can see in the cloud numbers and for a free cash flow trend that turns, not for a payout habit that has just been rewritten. Whether you should own them at all depends on your own situation and risk tolerance, and that is not something a column can settle.