Amazon’s free cash flow just swung $26 billion in the wrong direction, and the stock still ripped 15% higher. If you want to understand how markets really work in 2026, this Amazon free cash flow story is the perfect case study.

The company reported a trailing-12-month free cash flow outflow of $7.6 billion, compared with an $18.2 billion inflow a year earlier. On paper, that’s a glaring red flag. In practice, investors shrugged and pushed shares near the top of their 52-week range.

Amazon free cash flow vs. what really matters

Free cash flow is usually one of the cleanest ways to judge a business. It’s the cash left over after operating the company and funding the capital spending needed to keep it running. Negative free cash flow often means a business is burning cash, and that’s rarely a good headline.

But context is everything. Amazon’s operations aren’t deteriorating — they’re surging. Operating cash flow over the last 12 months rose 33% year over year to $161.4 billion. That’s the engine of the business, and it’s throwing off far more cash than it did a year ago.

The culprit behind the free cash flow plunge is capital spending. Purchases of property and equipment, net of proceeds, reached $169 billion over the same period, up 64% year over year. In the second quarter alone, Amazon poured $54.2 billion into capital expenditures, up from $32.2 billion in the prior-year quarter.

In other words, the negative free cash flow isn’t a sign of a broken business model. It’s the result of Amazon choosing to spend even faster than its cash engine is growing.

Where all that cash is going: AI and AWS

This is not a mystery capex binge. Amazon has been explicit: the lion’s share of that capital is going into artificial intelligence infrastructure, particularly inside Amazon Web Services (AWS).

AWS revenue jumped 37% year over year to $42.2 billion, marking its fastest growth in 18 quarters. For a cloud giant that was already massive, that reacceleration is hard to ignore. The company has also signaled that its capital spending plans are only getting larger, with expectations for roughly $220 billion of capex as it builds out AI-ready data centers and the hardware that powers them.

That’s why investors didn’t run when they saw the free cash flow number. They’re treating it less like a leak in the hull and more like a deliberate decision to floor the gas.

The capex story in one snapshot

Here’s how the key numbers stack up over the trailing 12 months:

  • Operating cash flow: $161.4 billion, up 33% year over year
  • Capital spending (property and equipment, net): $169 billion, up 64%
  • Free cash flow: negative $7.6 billion, down from positive $18.2 billion
  • AWS revenue: $42.2 billion, up 37%, fastest growth in 18 quarters

That is not what trouble looks like. It’s what a massive reinvestment cycle looks like.

Why the market cheered a negative free cash flow print

So why did Amazon stock jump about 15% on the heels of a quarter where free cash flow went negative?

First, investors increasingly separate why free cash flow is negative from the fact that it’s negative. There’s a huge difference between a company bleeding cash to stay alive and one choosing to spend aggressively into a high-return opportunity like AI infrastructure.

Second, Amazon’s results checked the other boxes that matter: operating cash flow climbing sharply, and AWS growth accelerating just as the AI spending wave is hitting full stride. That combination tells a simple story — the business is funding its own expansion, and the segment absorbing the cash is also the one growing the fastest.

Third, markets are looking forward, not backward. A trailing-12-month free cash flow number bakes in quarters when Amazon was spending heavily to lay groundwork for AI services that will show up in revenue and margins over the coming years, not just the last 12 months.

Investors are effectively saying they’re comfortable trading today’s free cash flow for tomorrow’s higher structural earnings power — especially in AWS.

Data center corridor illustrating Amazon free cash flow investment in AI infrastructure
Amazon’s negative free cash flow reflects a massive build-out of AI-ready cloud infrastructure. (Photo: Neil Palmer/CIAT / CC BY-SA 2.0 via Wikimedia Commons)

Amazon’s AI bet: risk, reward, and timing

None of this makes the free cash flow swing trivial. A $26 billion move in the wrong direction is real money, even for Amazon. The question is whether the return on that investment will justify the hit.

The company is directing tens of billions of dollars into data centers, networking gear, and specialized hardware designed for AI workloads. The goal is straightforward: make AWS the default place to train and run AI models, and lock in long-term, high-margin cloud revenue.

AWS already sits at the center of Amazon’s profit story, and the latest 37% revenue growth to $42.2 billion shows that demand for its services is not just holding up — it’s reaccelerating. That makes it easier for investors to stomach short-term free cash flow pressure. When the high-growth, high-margin segment is also the one absorbing most of the capex, the long-term math can look compelling.

The flip side: this is not a risk-free strategy. If AI spending across the industry cools, if customers push harder on pricing, or if competitors win more of the most lucrative workloads, Amazon could find itself with an AI infrastructure footprint that’s bigger, and more expensive, than it needs. In that scenario, today’s negative free cash flow would look a lot less like smart investing and a lot more like overbuilding.

How investors should read Amazon’s free cash flow now

For anyone watching Amazon from the sidelines, this quarter is a reminder that free cash flow headlines don’t always tell the full story. You need to follow the cash, not just the number.

Here are three questions to ask when you see a similar move from any large-cap tech name:

  • Is operating cash flow growing? In Amazon’s case, a 33% jump to $161.4 billion says the core business is strong.
  • What’s driving capital expenditures? Here, a 64% surge in net property and equipment spending to $169 billion is tied primarily to AI and cloud infrastructure.
  • Is the growth engine accelerating? AWS revenue up 37% to $42.2 billion — its fastest growth in 18 quarters — suggests the segment consuming the investment is also powering the story.

When those three answers line up, negative free cash flow starts to look less like a warning sign and more like a strategic choice.

What this signals for Big Tech and AI

Amazon is not the only giant throwing staggering amounts of cash at AI, but this quarter crystallizes the trade-off every large tech platform is making: near-term free cash flow versus long-term AI dominance.

For investors, the Amazon free cash flow swing underscores that traditional metrics are being warped by one of the largest infrastructure build-outs the industry has ever seen. Multi-billion-dollar data centers, specialized chips, and networking upgrades all hit cash flow statements before they fully show up in revenue.

That doesn’t mean free cash flow suddenly stops mattering. It means you have to separate cyclical cash burn from deliberate, front-loaded investment into assets that can earn outsized returns over years.

What This Means

Amazon’s $26 billion free cash flow swing is a stress test for how much investors are willing to tolerate in the name of AI-driven growth. For now, the answer is: quite a lot. The stock’s 15% surge tells you the market cares more about a 33% jump in operating cash flow and AWS revenue growing 37% than a single negative free cash flow print.

If AWS continues to scale into its new AI-heavy footprint, this quarter’s cash burn will look like the cost of securing a dominant position in the next phase of cloud computing. If the returns disappoint, that same number will become Exhibit A in the case against runaway capex.

Either way, this is the new reality for Big Tech: the biggest stories won’t just be about earnings per share. They’ll be about who’s willing to sacrifice tidy short-term free cash flow for a shot at owning the infrastructure of the AI era — and whether those bets actually pay off.

Photo: Astro_Alex / BY via Openverse | Photo: Neil Palmer/CIAT / CC BY-SA 2.0 via Wikimedia Commons