Meta's free cash flow went negative. Now it plans to sell stock.

Hey {{first_name|Investor}} -

Meta's ad business had one of its best quarters, and its cash flow had one of its worst. On Wednesday it guided to fourth-quarter spending that could push 2026 free cash flow below zero, and it filed a document saying it expects to sell stock to fund the buildout.

Last Sunday I listed 4 signs that a giant's borrowing starts to matter, and said Meta's report on October 21 was the first test. This is that test. I also ran the call through my earnings analyzer and put my own Meta thesis through a stress test, because I'd rather find the hole than have the market find it for me.

Revenue: $64.6 billion, up 26%. Ad impressions grew 10% and the average price per ad grew 14% (Meta Q3 release, October 21). Meta guided fourth-quarter revenue to $73 billion to $76 billion, which is 22% to 27% growth on last year's $59.9 billion.

Operating income: $13.0 billion, down 37%. That includes a $9.89 billion charge for the August agreement with 52 state attorneys general, which provides for payments of up to about $18.59 billion over 10 years (Meta Q3 10-Q). Without the charge, operating income was $22.9 billion, up 11% (Meta Q3 release).

Free cash flow: negative $2.1 billion. Cash from the business was $32.8 billion. Spending on data centers, servers and network gear, including finance-lease payments, was $34.9 billion (Meta Q3 release). A year ago, free cash flow for the same quarter was positive $10.6 billion.

Stock pay: $7.4 billion, up 33%. Headcount fell 15% over the same year, to 66,775 (Meta Q3 release). It's rough math on quarter-end headcount, but stock pay per employee went up a lot.

The ad business is fine. It's better than fine. The question is what it's being asked to pay for.

1. Free cash flow negative for a full year: a real risk. Meta raised the low end of its 2026 spending guide, to $135 billion to $145 billion including finance-lease payments. It spent $85.8 billion in the first 9 months. That leaves $49 billion to $59 billion for the fourth quarter, against $34.9 billion in the third.

Free cash flow for the first 9 months was $11.0 billion. To keep the full year positive, fourth-quarter cash from operations has to top about $38 billion at the low end of the spending guide, $43 billion at the midpoint, or $48 billion at the high end (my math on Meta's filings). Last year's fourth quarter, Meta's best in the last two years, brought in $36.2 billion. So the low end needs about 5% growth on that, which a business growing revenue 22% to 27% can clear. The midpoint needs about 19%, and the high end about 33%. Meta doesn't guide cash flow, so this is my read: the more of the spending guide Meta uses, the likelier a negative year.

2. Commitments outside the headline numbers: tripped, and growing. As of September 30, Meta had about $355.7 billion of leases signed but not yet started, mostly data centers, energy infrastructure and colocation, starting between now and 2036 (Meta Q3 10-Q). In June the figure was about $279 billion, before roughly $68 billion more was signed in July (Meta Q2 10-Q).

The guarantees grew too. Meta's maximum residual-value guarantees to the data-center ventures it doesn't consolidate, including its Louisiana and Texas campuses, rose to $41.5 billion from $28.1 billion at the end of 2025 (Meta Q3 10-Q). Meta would pay only if it walks away from a lease and the campus is worth less than the guaranteed amount. Meta says that isn't probable, so it has recorded no liability. These aren't bills coming due. They're commitments that don't show up as Meta debt.

3. Shareholder payouts shrink: still tripped. Meta has bought back no stock for 4 straight quarters. It paid $1.36 billion in dividends in a quarter when free cash flow was negative (Meta Q3 release; SEC filings).

4. Financing gets harder while coverage weakens: an early warning. Meta didn't issue bonds in the third quarter. Its $24.9 billion of net proceeds from new debt this year all came in the first half (Meta Q3 release). But the price of insuring Meta's debt against default, the 5-year credit default swap, hit a record 108 basis points in mid-October, meaning about $108,000 a year to insure $10 million of Meta bonds (Financial Times, October 20). Part of that is fund managers using these contracts as a cheap hedge on AI stocks generally, so I wouldn't read it as lenders doubting Meta. It's still the direction sign 4 warns about. And Meta is now planning to raise equity, which my 4 signs didn't cover. More in the next section.

So 2 of the 4 have tripped, a third is a real risk depending on how much of its spending guide Meta uses, and the fourth is an early warning. Last week I said one sign alone can be noise, and two or three together are worth paying attention to.

The rest of this issue is for Insiders.

Below the line: the same-day filing that says Meta expects to raise equity "in the near term," and what its new nonvoting shares mean for your slice of the profits. The Muse cost question Meta wouldn't answer on the call. The $25 billion New Mexico judgment. And the exit rule in my own Meta thesis that my stress test showed was broken.