Amazon Spent $220 Billion to Prove It Doesn't Need Your Cash Flow Anymore

Here's a fun exercise: tell a room full of value investors that a company just posted negative free cash flow over the trailing twelve months, and watch them reach for the exits. Now tell them that same company just printed the first $200-billion revenue quarter in corporate history, grew its cloud business at the fastest clip in 18 quarters, and has analysts tripping over each other to slap $325-335 price targets on the stock. Watch the room split right down the middle. That's Amazon right now, and honestly, that's the whole story.

The Quarter That Broke the Calculator

Back on July 30, Amazon dropped a Q2 report that did exactly what it needed to do: it shut up the skeptics and fed the believers in the same breath. Net sales hit $200.6 billion, up 20% year-over-year and comfortably past the ~$196.5B Wall Street was penciling in. AWS — the actual engine of this entire enterprise — grew roughly 37% to $42.2B, its best growth rate in a year and a half, with a $496 billion backlog sitting behind it like a loaded spring. Andy Jassy didn't even try to be humble about it, tossing around the phrase "$1 trillion annual revenue business" like it was a foregone conclusion rather than a moonshot.

Advertising, the quiet profit machine nobody talks about enough, grew 26% to $19.8B at margins north of 50%. AI and custom silicon each crossed a $25 billion annualized run rate at triple-digit growth. This is not a company running out of ideas.

The stock agreed, at least initially — up nearly 12% in the month following the print, briefly trading near $270.58. Then, as these things go, gravity remembered it exists and shares gave back roughly 7.5% of that move. Welcome to trading a $2 trillion-plus market cap company: even the good news comes with an asterisk.

The Asterisk, Specifically

Here's where I get less enthusiastic. That headline $5.75 EPS number was carrying some dead weight — about $1.2 billion in one-time tariff refunds and energy-contract gains, and as much as $53.4 billion in non-operating income padding the picture. Strip that out and the beat is still real, but it's not quite the clean knockout the headline suggested.

More importantly: Amazon raised its 2026 cash capex guidance from $200B to roughly $220B, with AWS alone potentially eating $165B of that, up more than 70% year-over-year. The bill for all this AI infrastructure ambition showed up exactly where you'd expect — trailing free cash flow flipped negative, to the tune of -$7.6 billion. That's not a rounding error. That's Amazon telling you, in plain arithmetic, that the AI buildout comes first and shareholder cash comes... later. Maybe much later.

Layer on continued job cuts as part of an "AI-focused cost reset," a reported $60 billion Qualcomm chip deal adding supply-chain complexity, and the fact that mega-caps like Amazon made up roughly 37% of S&P 500 market cap as of last October — meaning any broad multiple compression hits this stock disproportionately hard — and you've got a legitimate bear case that isn't just noise from people who missed the rally.

So What Do You Actually Do With This

Nothing dramatic, is the honest answer. This is a stock caught between a structurally excellent growth story and a self-inflicted cash flow crater, and until one of those forces clearly wins, the sensible move is to sit in the range and let the Q3 report — due October 29 — settle the argument. Support looks like it lives around $195-$210 if the FCF story spooks the market further; resistance sits at $270-$280, and punching through that needs guidance that actually eases the capex anxiety, not just another AWS growth headline.

The bulls are right that AWS demand is outstripping capacity through 2027 and beyond, per Jassy himself. The bears are right that spending $220 billion while burning cash is a bet, not a certainty. Both of these things are true simultaneously, which is exactly why this is a hold, not a hero trade.

Amazon isn't broken. It's just expensive to build the future, and right now the company is handing you the invoice before showing you the finished building. Come October 29, we find out whether construction's on schedule — or whether the contractor needs another change order.