To The Moon

$Amazon.com(AMZN)$  

It borrowed more than it burned.

Versus history: twelve-month free cash flow is now minus $42.4B. A year ago it was plus $14.8B. Long-term debt went from $65.6B at the end of December to $173.1B on 30 September. That is about $107.5B added in nine months.

First, Amazon's own explanation. The release says free cash flow fell "driven primarily by a year-over-year increase of $94.8 billion in purchases of property and equipment", and that this "primarily reflects investments in artificial intelligence". It also kept its guide of about $220B of cash capex for 2026. So the company's story is simple: we are building AI capacity, and AWS is growing 39% at a record 38.7% margin. That is the case it makes. Here is mine.

Yesterday I set Amazon two tests. Scoring them honestly.

Test 1. Does the funding gap close from operations, or from new money?

New money. Debt, and more than the gap. Hit.

Test 2. If Amazon's lease assets jump while capex lands near guide, Microsoft's lease shift is industry-wide.

New operating leases were $7.4B against $77.3B of cash capex, and the roughly $220B guide held. Miss. So far the lease shift looks like a Microsoft thing, not an industry thing.

One hit, one miss. I would rather show you the miss than quietly drop the test.

Now the part worth your time. Three cloud prints this week, lined up by who paid.

Microsoft. Azure up 48%. Capex paid from operating cash. Stock up.

Alphabet. Cloud up 103%. Buybacks cut to zero, $12.4B of stock sold. Stock down.

Amazon. AWS up 39% at a record 38.7% margin. $48.1B of new debt. Stock down.

Sort them by cloud growth and the reactions make no sense. Alphabet grew fastest and fell. Sort them by who paid and they line up.

My hypothesis, labelled as one: the market is now pricing who pays for the capex, not how fast the cloud grows. Three prints is not proof. It is a pattern worth testing.

How to run it on a company I never mentioned. The funding ladder.

1. The gap. Net capex minus operating cash flow. Amazon Q3: $75.9B minus $42.4B is $33.5B.

2. New debt. The financing section, proceeds from long-term debt. Amazon: $48.1B.

3. New shares. Stock sold, or buybacks quietly stopped. That was Alphabet's route. Amazon showed no stock sale line.

4. New leases. In the supplemental cash-flow table. Amazon: $7.4B operating, $1.5B finance.

5. Equipment received but not yet paid for. Same table, a line most people skip. Amazon, nine months: $21.5B, against $4.8B a year earlier. In effect, credit from suppliers.

Rung 5 is the one I would look at first on any big spender. Cash capex can look under control while the unpaid pile grows. For Amazon, Q3 alone added only $0.9B, so most of that build-up came earlier in the year. I have not measured how much of it gets paid in Q4.

A household version, made-up numbers. Say a family's renovation costs 100 units and its salary leaves 55 units spare. The gap is 45. You can fill it from a loan, from selling something, from a hire-purchase plan, or by telling the contractor you will pay next quarter. Same renovation, four very different risks. The ladder just asks which one they used.

The case against me.

Borrowing at investment-grade rates to fund a business growing 39% at a 38.7% margin can be cheap capital, not a warning. Corporate bond spreads barely moved: 0.80% in December, 0.86% on 28 October. Q3 also had about $24.0B of "acquisitions, non-marketable investments, and other", so not all of the new debt went into data centres. And the after-hours drop also reflects the earnings per share and the Q4 guide, not only funding. What would change my mind: Amazon funding Q4 without new debt, or Microsoft falling on a print while paying its own way.

Tests, with dates.

The 10-Q, due within days: what that $24.0B line actually is.

Nvidia, 17 November US time: if suppliers are extending credit, its receivables should grow faster than revenue.

Investment-grade spreads on 30 November: my hypothesis is still under 1.00%.

Which rung of the ladder do you check first when a company's capex runs ahead of its cash?

Not advice. I may hold positions in the names mentioned.

Stop guessing. Start calculating.

Live to fight another day. 🤙

#options #tradingpsychology #riskmanagement #optionstrading