ORLANDO, Florida, Nov 3 (Reuters) - The reaction of most

"Magnificent Seven" tech giants' shares to their latest earnings

suggests the artificial intelligence boom is far from over. Yet

doubts about the future returns from these firms' astronomical

AI expenditures are gnawing deeper.

The third quarter earnings season has seen these tech

behemoths continue to rake in huge profits and offer sunny

guidance. Some investors may baulk at the Mag 7's lofty

valuations, but today's tech leaders - unlike the superstar

firms of the 1990s dotcom bubble - appear to have sustainable

business models. Federal Reserve Chair Jerome Powell reiterated

as much last week, saying that their AI investments are a major

source of U.S. economic growth.

Just four "hyperscalers" alone - Microsoft ( MSFT ), Amazon ( AMZN )

, Meta and Alphabet - are expected

to spend a combined $350 billion this year, and Goldman Sachs

estimates global AI-related infrastructure spending could reach

$4 trillion by 2030.

The more these firms splurge on data centers, cloud

computing capabilities, and the gamut of AI technologies, the

loftier investors' expectations will get. At some point, they

will be impossible to meet.

The financial benefits and cost savings for society

resulting from that are one thing; which companies actually

profit is another. It is important, therefore, to distinguish

between "value creation" and "value capture".

"The value creation is certainly there," says Daniel Keum,

associate professor of management at Columbia Business School.

"But will that value flow back to the companies that are making

these AI investments right now? For me, the clear answer is no."

DO THE MATH

It's early days in the AI supercycle, but Big Tech's AI

outlays are already eating into hyperscalers' cash flows.

Torsten Slok, chief economist at Apollo Global Management,

estimates that aggregate capex at Amazon ( AMZN ), Google, Microsoft ( MSFT ),

Meta and Oracle as a share of their operating cash flow is now a

record 60% - and rising.

Amazon ( AMZN ) reported strong earnings last week, and its stock

surged double digits to hit a new high on Friday. But buried in

the report was a slide showing that trailing-12-month free cash

flow has fallen almost 70% over the last year.

Ross Hendricks, analyst at independent research firm Porter

& Co, estimates that hyperscalers' free cash flow in the first

quarter of next year will be down more than 40% from the same

period this year.

"The whole sector faces the same basic problem," says Bob

Elliott, co-founder of Unlimited Funds. "The math is pretty

simple, unless there is a surge in revenues from these

activities, Big Tech is going to pump nearly all their free cash

flow into capex in just a few years."

This creates several potential problems. It intensifies the

pressure to generate high returns on these investments, but

until those materialize, non-AI-related activities are also

under pressure to produce significant returns. And this leaves

hyperscalers vulnerable in the event of a sharp economic or

market downturn.

HIGHER BAR

The fate of these megacaps will, of course, have a

significant impact on the broader economy, not only because

these companies' capex is helping to drive growth but also

because almost everyone with a retirement fund is exposed to

them. Nvidia's share of the total S&P 500 market cap is a

stunning 8%, while that of the "Mag 7" is a record 37%.

Investors are well aware of how much these shares have

appreciated. The Philadelphia Semiconductor Index has more than

doubled from its April low. But expensive markets can always get

more expensive.

It will take a brave fund manager to tell clients that

they're reducing exposure to what have effectively become

cash-printing machines. Of course, whether these companies can

continue printing money as fast as they're spending it is the

big question.

For example, Meta's announced capex this year is around $70

billion, but Unlimited Funds' Elliott notes that the company's

income is only $3 billion to $5 billion higher, based on

underlying trends, than it was before they started spending all

this cash. That's a pretty "mediocre" return on investment.

Of course, CEO Mark Zuckerberg might argue that this is

long-term investment and that not spending now could be more

costly down the line if the AI revolution lives up to the hype.

But it is unclear how much patience investors will have.

Smaller businesses overall seem to be faring better. A

Wharton Business School study published last month found that

74% of businesses say generative AI investment is already

producing positive returns, especially smaller enterprises in

digital-based sectors like tech and finance.

"Confidence remains strong ... but future gains must now be

justified by clear performance outcomes," the authors said.

The bar for Big Tech giants with market caps of trillions of

dollars and capex budgets of hundreds of billions is higher

though. Much higher.

(The opinions expressed here are those of the author, a

columnist for Reuters)

Enjoying this column? Check out Reuters Open Interest (ROI),

your essential source for global financial commentary. ROI

delivers thought-provoking, data-driven analysis of everything

from swap rates to soybeans. Markets are moving faster than

ever. ROI can help you keep up. Follow ROI on LinkedIn and X.