China’s economy slows further in July as retail sales barely grow

Big Tech’s record AI spending is starting to look less like a drag on valuations and more like the foundation for the next leg of earnings growth.

Results from Microsoft, Amazon, Alphabet and Meta offered fresh evidence that demand is catching up with the infrastructure build-out.

Azure grew 43%, AWS 37% and Google Cloud 82% in their latest quarters, while Meta’s revenue rose 28%.

That matters because investors have spent much of 2026 questioning whether hundreds of billions of dollars in data-centre spending would translate into durable returns.

Capex is still enormous, but the payoff is becoming visible

Microsoft, Amazon, Alphabet and Meta are on track to spend a combined $730 billion to $760 billion in 2026, according to Janus Henderson, with roughly $430 billion expected in the second half.

For months, that spending looked like a reason to mark the stocks down.

Free cash flow has come under pressure, and the hyperscalers have lagged the semiconductor rally despite writing the biggest cheques.

But the latest numbers have changed the debate. Microsoft’s commercial remaining performance obligations reached $678 billion, while Google Cloud’s backlog rose to $514 billion.

Janus Henderson estimates the three biggest cloud platforms now have more than $1.6 trillion of contracted backlog.

Wellington Management’s Brian Barbetta sees hyperscalers as likely to emerge as major beneficiaries of the AI shift, and the firm has raised exposure to several of them.

The cloud companies may be entering the monetisation phase

The investment case is no longer simply that AI demand is strong. It is that capacity brought online today can turn into higher revenue over the next several years.

Microsoft said customer demand still exceeds available Azure capacity.

Alphabet raised its 2026 capex forecast to $195 billion-$205 billion because demand is arriving faster than expected. AWS growth accelerated to 37% as more AI workloads moved into production.

Richard Clode of Janus Henderson expects profit and cash-flow growth at hyperscalers to begin outpacing incremental capex growth by late 2027 into 2028.

That would mark an important turning point: spending that has compressed free cash flow would start working harder for shareholders.

Scale could separate the winners from the AI trade

The risks have not disappeared. Alphabet reported negative free cash flow of $5.9 billion in the second quarter, while Meta generated just $784 million as infrastructure spending surged. Meta expects 2026 capex of $130 billion-$145 billion.

That is why the next phase of the AI trade may favour companies with scale, deep customer relationships and control over their infrastructure.

Capital Group’s John Lamb argues investors should not treat chipmakers and hyperscalers as an either-or choice.

Both can benefit as spending moves through the AI ecosystem, but their return profiles will differ as new capacity starts generating revenue.

The market still needs proof that returns can justify the sums involved.

But with cloud growth accelerating and backlogs swelling, Big Tech’s AI capex is beginning to look less like the problem investors feared, and more like a reason earnings could keep growing.