S&P Global Ratings revised its outlook on Meta Platforms to negative from stable on October 30, 2026, and affirmed the company's AA- issuer credit rating, warning that leases signed for giant AI data centers could push Meta's debt measure above the agency's downgrade threshold as early as 2027. The ratings agency now forecasts Meta's capital expenditure at $140 billion in 2026, more than $225 billion in 2027 and $252 billion in 2028, against earlier forecasts of $163 billion and $178 billion for the last two years. It expects free operating cash flow deficits of $22 billion in 2026, $71 billion in 2027 and $39 billion in 2028, where it had previously expected Meta to return to positive cash flow in 2028. The report also casts doubt on how quickly Meta can earn back the money: "In our view, Meta's ability to monetize its AI investments is more limited than that of its peers."

In Short

A credit rating agency, S&P, still gives Meta one of its best grades, but it has warned that the grade could fall because Meta is spending far more on AI data centers than it earns in cash. Most of the risk comes from long leases on new data centers, which S&P counts like debt, and from the fact that Meta does not yet make clear money from AI outside its ads. If Meta's spending stays this high and products such as its Muse assistant do not start paying their way by 2028, S&P says it could lower the rating.

What S&P changed

The rating action was published from New York on October 30, 2026, under the headline "Meta Platforms Inc. Outlook Revised To Negative On Significant Uncertainty Around Capex And Free Operating Cash Flow." Naveen Sarma, based in New York, is the primary analyst. Jawad Hussain in Chicago and David T Tsui in San Francisco are listed as secondary contacts.

Two things happened at once. The AA- rating, the fourth-highest grade on S&P's scale, was affirmed, so Meta's credit standing has not been lowered. The outlook, which signals the direction S&P thinks the rating may move, was moved to negative from stable. According to the report, the change "reflects our expectation that the company's leverage could exceed 1.0x and free operating cash flow (FOCF) could remain negative for longer than expected if capital expenditure (capex) remains elevated and the monetization of its AI investments is delayed or fails to materialize."

A 1.0x threshold is tight, and it reflects how little debt Meta has carried. S&P puts the company's adjusted leverage at about 0.1x in 2025. It now expects that ratio to exceed 1.0x, the level at which an AA- rating comes under pressure, as early as 2027. S&P calls this a "dramatic increase."

A capex forecast raised by at least $62 billion

The core of the decision is a revised view of what Meta will spend. Management updated its 2026 capital expenditure guidance on October 21 to between $135 billion and $145 billion, a $5 billion increase at the low end from the previous quarter, as PPC Land reported with the third-quarter results. S&P places its 2026 forecast at $140 billion, the midpoint of that range.

The later years moved far more.

Year S&P forecast now S&P forecast before Change (PPC Land calculation)
2026 $140 billion not stated -
2027 over $225 billion $163 billion at least +$62 billion, about +38%
2028 $252 billion $178 billion +$74 billion, about +42%

By PPC Land's calculation, the 2027 figure is at least 61% above the 2026 forecast. Growth then slows to about 12% in 2028, which is what S&P means when it says it forecasts "capex growth flattening out beginning in 2028." Across the three years, the forecasts add up to more than $617 billion.

Why the jump? S&P points to the absence of information as much as to anything Meta has said. "Given the lack of clarity around the company's future capex (it has yet to disclose its spending plans for 2027) and its very ambitious multi-year AI investment plans, we are less certain that it will curtail its spending as previously anticipated," the report states. On the October 21 call, chief financial officer Susan Li said Meta was at an early stage of its budget and capacity planning for 2027 and would give specific guidance on the fourth-quarter call.

The scale is clearer in comparison. Meta spent $85.82 billion on capital expenditure, including principal payments on finance leases, in the first nine months of 2026. Reaching S&P's $140 billion would mean about $54.2 billion in the fourth quarter alone, PPC Land's calculation, more than any quarter Meta has reported. And the forecast of more than $225 billion for 2027 would exceed the roughly $220 billion of cash capital expenditure Amazon expects for all of 2026, as well as the $195 billion to $205 billion range Alphabet set for 2026 in July.

Three years of cash deficits

Capital spending on that scale outruns the cash Meta's apps produce. S&P now expects these free operating cash flow figures:

Year S&P forecast
2026 -$22 billion
2027 -$71 billion, which S&P calls "the high water mark"
2028 -$39 billion

That is a cumulative shortfall of $132 billion over three years, by PPC Land's calculation. The previous forecast had Meta generating positive free operating cash flow again in 2028.

The third quarter was the first in which the deficit became visible. According to S&P, Meta generated "a FOCF deficit ($2.1 billion) for the first time during the quarter because its capex, which includes principal payments on finance leases, increased to $34.9 billion." Meta's own figures match. Its free cash flow was $12.39 billion in the first quarter, $784 million in the second and negative $2.12 billion in the third, as its results presentation showed. That is a positive $11.05 billion over nine months, by PPC Land's calculation.

That leaves a gap the report does not explain. If S&P's measure works the same way for the full year as it does for the third quarter, a $22 billion deficit for 2026 implies a fourth-quarter shortfall of about $33 billion, PPC Land's calculation. S&P does not publish a quarterly path, and its annual figure may include items that Meta's quarterly reconciliation does not.

Leases, not bonds

The report is explicit about where the leverage comes from. The increase "primarily stems from its strategy of leasing data center capacity (including its mega-data center projects Beignet and Sopaipilla) rather than from issuing debt." S&P counts those lease obligations in its adjusted debt figure. So a company can raise its leverage on S&P's measure without selling a single bond.

The numbers behind that are in Meta's quarterly filing for the period ended September 30, 2026. According to S&P, Meta's future lease obligations rose to $355.7 billion from approximately $279 billion in the second quarter, an increase of about $76.7 billion, or roughly 27%, in three months by PPC Land's calculation. PPC Land reported on October 24 that the filing shows about $355.73 billion of signed operating and finance leases that had not yet started, within total non-cancelable commitments of $454.52 billion.

S&P identifies Sopaipilla as the proposed data center project in El Paso, Texas, and says the updated lease figure includes it. Meta disclosed a venture with BlackRock to develop a 1-gigawatt data center in El Paso on July 28, 2026. The same day, S&P published a bulletin titled "Meta Platforms Inc.'s Data Center Joint Venture Does Not Immediately Affect Ratings." Three months later, the effect has arrived.

"Our treatment of this transaction is identical to the approach we took with Meta's Louisiana joint venture with Blue Owl," the report says. Meta announced the Blue Owl joint venture for data center development in Louisiana with its third-quarter 2025 results. According to the quarterly filing, as reported by PPC Land, Meta holds 20% of each venture. The Louisiana venture has an estimated development cost of about $27 billion, with Meta's leases starting in 2029, and the Texas venture about $14 billion, with leases starting in 2028. Meta's maximum exposure to loss across the two is $68.40 billion, of which $41.49 billion is residual value guarantees.

The 2028 start in Texas matters for the forecast. "It is principally due to these lease obligations that we estimate the company's S&P Global Ratings-adjusted leverage will increase to 1.3x when the leases commence in 2028, which exceeds our 1.0x downgrade threshold for the current rating," the report states.

$37.5 billion of new equity, and $60 billion more planned

Meta has already moved to raise money without borrowing. According to S&P, the company "recently announced that it raised $37.5 billion by issuing $25.0 billion of Class A common stock and $12.5 billion of Series A mandatory convertible preferred stock." It also plans to raise a further $60 billion by selling Class C stock through an at-the-market program starting in January 2027.

The timing stands out. On the October 21 call, Li said Meta had no specific financing plans to announce at that time, as PPC Land reported. The non-voting Class C shares that the at-the-market program would sell do not yet exist: the board approved them with the third-quarter results, effective January 19, 2027, and named capital raising as one of their purposes.

S&P calls the raise "credit positive" because Meta "could use this cash to cover its forecast FOCF deficit for 2027 instead of debt, as we had previously projected." Its leverage path shows how much difference the equity makes, and where it stops helping:

Year-end Earlier S&P forecast New forecast, $37.5 billion raise included New forecast, if the $60 billion program is also included
2026 0.4x 0.3x -
2027 0.7x 1.1x 0.7x
2028 0.8x 1.3x 1.1x

S&P's base case leaves out the $60 billion program. "Note that we do not include future equity issuance in our forecast," the report says. Even with it, the 2028 figure of 1.1x sits above the 1.0x threshold. S&P also expects Meta "will solely use debt to cover its expected FOCF deficit in 2028 because it is unclear if it plans to issue more equity to fund its future cash needs."

Meta is not the first large technology company to turn to share sales this year. In June, Alphabet announced about $85 billion of equity financing, including a $10 billion private placement with Berkshire Hathaway and a $40 billion at-the-market program. By the end of September, Alphabet's free cash flow had turned negative for a second consecutive quarter, at minus $9.1 billion.

The third quarter through S&P's eyes

None of this reflects weakness in the advertising business, according to S&P. "Meta Platforms' third-quarter 2026 results demonstrated the strength of its core business," the report says. Revenue rose 26.1%, led by a 25.9% increase in advertising revenue, with double-digit growth in both ad impressions and average price per ad. The results showed impressions up 10% and average price per ad up 14%, as PPC Land reported.

Costs rose faster. S&P attributes part of that to the $9.89 billion charge tied to the settlement with state attorneys general over teen mental health, which "resulted in a 2,000 basis point decline in its reported operating margins." The operating margin fell to 20% from 40% a year earlier. Meta now expects total expenses of $175 billion to $179 billion for 2026, up from the $165 billion to $169 billion range it gave in the second quarter, a revision S&P says reflects the charge.

S&P's description of the franchise remains strong. It holds a "strongly favorable view of the company's business position," based on "market-leading social media and messaging platforms, large and engaged global user and advertiser base, and significant proprietary first-party user data, which supports its advertisers' ad targeting and conversion." It calls Meta "the second-largest digital advertising company in the world after Alphabet Inc." and counts 3.64 billion engaged users, the same figure Meta reports as daily active people. S&P also sees room to grow through monetization of Messenger and WhatsApp and of features in Facebook and Instagram "such as Reels."

The monetization question

The difficulty, as S&P sees it, is not whether Meta's AI spending improves its products but whether it produces enough new revenue, soon enough, to cover its cost. The company intends to monetize the investment "via improved advertising tools that support faster ad revenue growth, as well as from the development of new revenue streams stemming from its introduction of agentic apps such as Muse," the report says. Yet "it is unclear when these revenue streams will generate material contributions that exceed their rising costs over the next few years."

The contrast S&P draws is with companies that sell computing capacity. "Several of the company's peers are already monetizing their data center investments by leasing compute capacity to third parties, including frontier model labs and enterprise customers." Meta reports two segments, Family of Apps and Reality Labs, and has no cloud business that rents capacity to outside customers. Its data centers mainly serve its own apps and models; the enterprise offering Zuckerberg described on October 21 is a model API, not rented capacity.

The peers' customer contracts are large. Anthropic's IPO filing on October 26 showed minimum compute commitments of at least $114.5 billion with Google Cloud, at least $110.0 billion with Amazon and at least $44.9 billion with Microsoft, a total of $269.4 billion by PPC Land's calculation. Amazon Web Services revenue rose 39% to $45.98 billion in the third quarter, PPC Land reported, and Google Cloud revenue rose 103% to $30.71 billion. Concentration has its own risks: UBS estimates cited by Ed Zitron in July put OpenAI and Anthropic at more than 48% of Google Cloud revenue in 2027, a projection rather than a reported figure.

Muse is the new revenue source S&P names. It describes Meta's "recently released personal agentic tool" as something that "could act as a significant source of growth to expand its business beyond its core digital advertising focus." Muse went live in the United States on September 8, and on September 29 Meta gave a small business version access to Meta ad accounts and 15 third-party tools. On the October 21 call, Mark Zuckerberg said more than 10 million people in the US and Canada had set up Muse agents and that Meta planned to charge businesses a fee on transactions Muse helps complete. No fee level or start date was given. Li declined to give figures on the cost of running Muse, and said Meta expected new AI businesses to begin contributing revenue in 2027 and to scale over time.

Muse is one of the agentic AI products S&P is weighing against more than $600 billion of forecast capital spending. Whether a transaction fee on shopping, subscriptions for advanced models and a business tier can generate cash on that scale by 2028 is the question the rating now turns on.

What would move the rating

The report sets out both directions. S&P could lower the rating if it expects adjusted leverage to rise above 1x "on a sustained basis" and free operating cash flow to remain negative beyond 2028. A second trigger concerns strategy rather than arithmetic: "we could lower the rating if we take a negative view on Meta's AI strategy due to ongoing uncertainty about the monetization of its AI investments."

A return to a stable outlook would require S&P to expect leverage to remain below 1.0x and to believe that Meta will generate positive free operating cash flow in 2028. According to the report, "This could occur if Meta shifts its capex spending, increases the monetization of its AI investments, or raises additional capital without increasing its outstanding debt."

All three routes are visible already. Meta controls its spending plans for 2027, which it has not yet published. The at-the-market program is equity, not debt. And monetization depends on the advertising tools and Muse revenue that S&P says are not yet proven at the required scale.

Where this sits in S&P's coverage of AI spending

The report lists four earlier S&P publications that trace the same concern: a bulletin on Meta's data center joint venture on July 28, 2026; a tear sheet titled "Meta Platforms Inc.'s Increased Future Leases Are Driving Leverage Higher" on August 6; slides on S&P's view of artificial intelligence and hyperscalers on August 17; and "Credit Outlook For Hyperscalers: A Temperature Check" on September 3. The report lists 11 criteria documents behind the action, from S&P's corporate methodology to its principles of credit ratings.

S&P has already acted on another company over AI data center leases. On July 9, 2026, it lowered Oracle's long-term issuer credit rating to BBB- from BBB, the lowest investment-grade rating, after forecasting a free operating cash flow deficit of nearly $42 billion for Oracle's fiscal 2027 and adding $260 billion of lease commitments to its analysis. Oracle's downgrade threshold was leverage of 4x. Meta's is 1.0x. The two companies sit at very different points on the scale, but the mechanism S&P describes is the same: data center capacity contracted through long leases, counted as debt, ahead of the revenue meant to pay for it.

Meta belongs to the wider group of hyperscalers building AI capacity, and it is not alone in spending more than it generates. What separates it in S&P's analysis is the absence of an outside customer base for that capacity.

Why this matters for marketers

Meta's advertising business pays for this program. S&P's own figures show it: a 25.9% rise in advertising revenue in the third quarter was not enough to keep free operating cash flow positive once capital spending reached $34.9 billion in a quarter. The agency's path back to a stable outlook includes faster monetization of AI investment, and the first route S&P names for that monetization is "improved advertising tools that support faster ad revenue growth."

Advertisers already supply that revenue at rising prices. Average price per ad rose 14% worldwide and 23% in the US and Canada in the third quarter, while impressions grew 10%. The report does not discuss pricing, and it makes no forecast for ad rates. It does make clear that, in S&P's view, the near-term revenue comes from the existing advertising business, and that new streams such as Muse transaction fees remain a question of timing.

There is a second consequence for businesses that work with Meta's AI products. Muse's business model, as Meta described it on October 21, relies largely on fees paid by businesses on transactions, alongside subscriptions. S&P has now tied Meta's credit outlook partly to how quickly that kind of revenue appears, and its tests run to 2028.

Timeline

Summary

Who: S&P Global Ratings, with Naveen Sarma as primary analyst, and Meta Platforms, owner of Facebook, Instagram, Messenger and WhatsApp. The decision concerns Meta's creditors and shareholders, and indirectly the advertisers whose spending funds Meta's AI program.

What: S&P revised Meta's outlook to negative from stable and affirmed its AA- rating. It now forecasts capital expenditure of more than $225 billion in 2027 and $252 billion in 2028, free operating cash flow deficits of $22 billion, $71 billion and $39 billion in 2026 to 2028, and adjusted leverage of 1.1x in 2027 and 1.3x in 2028, above its 1.0x downgrade threshold, driven mainly by data center leases including the El Paso project it calls Sopaipilla.

When: The rating action was published on October 30, 2026. S&P's tests run to 2028, when the El Paso leases begin; Meta's $60 billion at-the-market equity program is planned from January 2027.

Where: The report was issued from New York. The leases in question cover data center projects in El Paso, Texas, and the Louisiana joint venture with Blue Owl.

Why: Meta's AI spending is rising faster than its cash generation, and unlike peers that rent capacity to AI labs and enterprises, it relies on better advertising tools and new products such as Muse to earn the money back. S&P says it is unclear when those revenue streams will exceed their rising costs, and that it could lower the rating if leverage stays above 1x and cash flow remains negative beyond 2028.