A negative outlook is a small thing in the language of ratings: the grade stays, the direction changes. For Meta Platforms, which holds one of the highest ratings in the S&P 500 and borrowed $30 billion in the bond market a year ago, the direction is the part that matters.
S&P Global Ratings revised the outlook on Meta’s AA- rating to negative from stable on October 30. The rating itself was affirmed. The agency cited capital spending and free cash flow, saying it was uncertain when the company’s AI investment would produce growth exceeding its cost, and that Meta’s path to monetization was less clear than that of peers.
A Meta spokesperson said the company disagreed with the assessment and described its credit position as strong.
The figures behind the revision are not ambiguous. Meta has guided to capital spending of as much as $145 billion this year. In the third quarter, free cash flow was negative $2.1 billion. A company generating negative free cash flow at that scale is funding its buildout from the balance sheet, and eventually from the bond market.
Meta has done both. It sold $30 billion of bonds in a single offering, among the largest corporate deals on record, and it has used structured financings for data centers, including a joint venture with Blue Owl Capital for a campus in Louisiana that raised more than $27 billion. The structures keep some of the debt off the headline balance sheet, but the obligation to pay for capacity remains.
The rating agencies look through the structure. S&P’s analysis of a hyperscaler weighs committed spending against expected cash generation, and Meta’s committed spending has risen faster than its cash generation for several consecutive quarters. Advertising revenue is still growing, and the core business remains one of the most profitable in the world. The question the agency raised concerns the destination of that profit.
Comparisons with peers cut both ways. Microsoft, Amazon and Alphabet are spending at similar or greater scale, and each has a cloud business that sells the capacity it builds. Meta builds for its own models and its own advertising systems. There is no third-party customer to absorb excess capacity if internal demand arrives late.
That is the substance of the concern about monetization. A cloud provider that overbuilds can rent the surplus. A company that builds only for itself has to convert the capacity into better ad targeting, more engagement or a new product line, and the conversion has to happen before depreciation eats the return.
For the bond market, the practical effect of a negative outlook is a wider spread on new issues and closer attention from rating committees at the next review. A downgrade to A+ would still leave Meta among the strongest credits in corporate America. It would also push the company below the threshold some conservative mandates use when they treat a rating as a proxy for safety, which matters for the cost of the next large deal rather than for this one.
The revision lands in a broader shift. The largest technology companies spent a decade funding expansion from operating cash flow, and their ratings reflected that. Over the past two years they have become among the largest borrowers in the investment-grade market, financing campuses, power contracts and processors with debt. The agencies are marking that transition one issuer at a time.
Meta is an unusual case because its rating was earned on a business with almost no capital intensity. Advertising software required servers, but nothing like the facilities now under construction in Louisiana, Ohio and elsewhere, each with dedicated power generation and cooling designed for racks that draw far more than a conventional data hall.
The company’s argument is that the spending is an investment in the same business. Better models produce better ad ranking, which produces more revenue per impression, which pays for the capacity. Management has pointed to engagement gains and to the early performance of AI-generated creative tools as evidence the returns are arriving.
S&P’s response is that the evidence is not yet sufficient to support the rating, which is a different claim. An outlook revision is a statement about the next two years, not about solvency. It tells investors that the agency sees a plausible path to a lower grade and wants the company to prove otherwise.
What would restore a stable outlook is not complicated to describe. Positive free cash flow, capital spending that grows more slowly than revenue, and disclosure connecting AI investment to specific revenue lines would each help. None is likely in the next quarter, given the construction already underway.
The company has options that a smaller borrower would not. It could slow the pace of new campuses, sell stakes in finished ones to infrastructure funds, or fund more of the program from operations as advertising grows. Each choice trades growth for balance-sheet strength, and the trade is easier to describe than to make when competitors are still building.
Meta can fund itself without the rating. It has the cash flow, the balance sheet and the market access. The rating measures how the market prices that access, and the measure has started to move.
