U.S. stock futures slipped on October 22, 2026, as two of the market’s biggest AI spenders delivered a split verdict on where the artificial intelligence buildout is headed. Nasdaq-100 futures fell 0.7% and S&P 500 futures dropped 0.4% after Tesla and Meta Platforms both posted third-quarter results that moved their stocks in opposite directions overnight, according to Investor’s Business Daily. The divergence centers on one line item: capital expenditure, the hardware and data center spending that underpins the entire AI trade.
Tesla cut its capital-spending outlook. Meta raised its own. The market punished the former and rewarded the latter, and the gap between those two reactions is now rippling through futures markets as traders try to figure out whether AI infrastructure spending is cooling off or just getting more selective. Yahoo Finance and other outlets flagged the earnings as the dominant story of the morning, alongside continued chatter about a possible Tesla-SpaceX combination.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
Tesla’s Third Quarter: Earnings Beat, Capex Pullback
Tesla reported third-quarter adjusted earnings of $0.48 per share on revenue of $28.05 billion. Those numbers landed as the headline figure of the report, but investors fixated on guidance instead. Tesla told investors it now expects capital expenditures of approximately $20 billion for 2026, a figure that triggered a 3.3% drop in Tesla shares in post-market trading once the capex cut became clear.
That reaction is notable because Tesla’s quarterly revenue and per-share profit were not the problem. The stock did not fall because the company missed on sales or margins. It fell because Wall Street reads capital spending as a forward indicator, and a smaller 2026 capex number signals less spending on the data centers, chips, and compute infrastructure that back Tesla’s self-driving and robotics ambitions. For a company that has spent two years telling investors its future value sits in AI and compute, trimming the hardware budget sends a specific signal, regardless of how the current quarter’s income statement looks.
Tesla’s Q3 2026 print follows a run of closely watched disclosures this year on vehicle deliveries and energy storage output, detailed in Tesla’s Q3 Earnings: 486,532 EVs, 13.7 GWh Storage. Those operational metrics were solid. The capex line is the one that moved the stock this time.
Why the Capex Cut Matters More Than the Earnings Beat
Capital expenditure guidance has become the single most scrutinized number in AI-adjacent earnings reports. It is the clearest proxy the market has for how much a company believes it needs in GPUs, custom silicon, data center buildings, and power infrastructure to support its AI roadmap. When that number drops, even modestly, it can be read as a company either pulling back on ambition, finding efficiencies that let it spend less for the same output, or quietly admitting that previous spending plans were too aggressive.
For Tesla specifically, capex underwrites the compute clusters behind Full Self-Driving training, the Dojo-adjacent infrastructure, and the hardware push behind Optimus. A cut to that budget, even a modest one, raises questions about pacing on all three fronts. Reports have not detailed exactly which programs absorb the reduction, and this story stays with what is confirmed: the capex figure dropped to roughly $20 billion and the stock fell 3.3% in response.
Meta’s Third Quarter: Revenue Jumps 26%, Capex Climbs Higher
Meta Platforms moved in the opposite direction on every metric that mattered to the market. The company reported third-quarter earnings of $3.83 per share on revenue of $64.6 billion, up 26% year over year. Meta shares rose more than 2% in extended trading once the numbers crossed the wire, a reaction that reflects both the strength of the top-line growth and the market’s apparent comfort with what Meta did next: it raised its own 2026 capital-expenditure outlook.
Meta now guides 2026 capex to a range of $135 billion to $145 billion, up from a prior lower-end figure of $130 billion. That is a meaningful upward revision on an already enormous base, and it came with no stock penalty. If anything, the combination of 26% revenue growth and a bigger hardware budget read as confirmation that Meta’s AI infrastructure spending is translating into measurable business results, not spending for its own sake.
The contrast with Tesla is stark. Two companies, both heavy AI infrastructure spenders, reported in the same 24-hour window. One cut its hardware budget and got sold off. The other raised its hardware budget and got bought. The market is not treating capex as inherently good or bad; it is treating capex changes as a referendum on whether the underlying business can justify the number.
Tesla vs. Meta: Side-by-Side Earnings Comparison
The table below lines up the confirmed figures from both reports, isolating the metrics that moved futures and after-hours pricing.
| Metric | Tesla | Meta Platforms |
|---|---|---|
| Q3 2026 adjusted EPS | $0.48 | $3.83 |
| Q3 2026 revenue | $28.05 billion | $64.6 billion |
| Revenue growth (YoY) | Not specified in guidance commentary | 26% |
| 2026 capex guidance | ~$20 billion | $135 billionâ$145 billion |
| Prior capex reference point | Higher than current guide (cut confirmed) | $130 billion (prior lower bound) |
| After-hours/post-market stock move | -3.3% | +2% or more |
The scale difference alone is worth pausing on. Meta’s 2026 hardware budget is now roughly seven times the size of Tesla’s, even though Tesla’s quarterly revenue is less than half of Meta’s. That is less a statement about which company is bigger and more a statement about how differently the two define their AI infrastructure needs. Meta runs the compute behind its recommendation engines, ad targeting, and consumer AI products at a scale that depends on continuous, massive GPU and data center investment. Tesla’s compute needs, while large, are tied more narrowly to autonomy training and robotics, and a smaller percentage cut there still shows up as the dominant story in its earnings reaction.
Futures and the Broader Market Reaction
The index-level reaction tells its own story about how the market is digesting mixed signals from its two biggest AI capex bellwethers in one session. Swissinfo described the broader pattern as the AI trade stumbling even as individual earnings beat expectations, a framing that fits what happened with Tesla and Meta specifically.
| Index/Asset | Move | Context |
|---|---|---|
| Nasdaq-100 futures | -0.7% | Tech-heavy index most exposed to AI capex sentiment |
| S&P 500 futures | -0.4% | Broader market, smaller but still negative move |
| Tesla shares (post-market) | -3.3% | Reaction to lower 2026 capex guidance |
| Meta shares (extended trading) | +2%+ | Reaction to 26% revenue growth and raised capex guidance |
Futures weakness across both major indexes, even as one of the two headline earnings reports beat on revenue growth, suggests the market is not simply rewarding growth. It is weighing capex direction as a signal about the durability of AI-related spending overall, and a cut from a company as closely associated with AI compute as Tesla is enough to drag sentiment down even when Meta’s number went the other way.
The Hardware Angle: What Capex Guidance Means for Chips and Data Centers
Capital expenditure at this scale is overwhelmingly a hardware story. It covers GPU and accelerator purchases, custom silicon contracts, data center construction, networking gear, and the power infrastructure needed to run all of it. Meta’s jump to a $135 billion-to-$145 billion range for 2026 represents one of the largest single-year hardware commitments disclosed by any public company, and it lands at a moment when the broader memory and storage market is already under strain. DDR5 pricing has climbed sharply this year as AI data center demand competes with consumer PC builders for the same supply, a dynamic covered in DDR5 RAM Hits $450 for 32GB as AI Shortage Bites.
GPU scarcity has followed a similar trajectory. High-end cards that would normally sell near list price have traded for multiples of their suggested retail price as data center buyers absorb available supply, a trend detailed in RTX 5090 Shortage: Prices Hit $8,500, 5x MSRP. A $10 billion-plus increase to Meta’s annual hardware budget, layered on top of already-constrained GPU and memory supply chains, points toward continued upward pressure on component pricing well into 2027, even as Tesla’s smaller pullback suggests at least one major buyer is tapping the brakes.
The capex question has also become central to how analysts evaluate Nvidia’s position, since both Meta and Tesla ultimately funnel large shares of their hardware budgets toward GPU and accelerator purchases. The circular nature of AI infrastructure financing, where chipmakers invest in customers who then buy more chips, has drawn its own scrutiny, as explored in Nvidia’s $500B AI Bet Splits From Its Tech Case.
Historical Context: How We Got to Capex as the Key Metric
Capital expenditure did not used to be the headline number in big tech earnings calls. Through most of the 2010s, investors focused on revenue growth, user counts, and margin expansion. That changed once generative AI shifted the center of gravity toward compute-intensive training and inference workloads. Starting in 2023 and accelerating through 2024 and 2025, capex guidance became the number analysts asked about first on earnings calls, because it is the clearest forward signal of how much a company believes AI will cost to build and how much revenue it expects that spending to eventually justify.
Meta has been one of the most aggressive capex spenders in that period, consistently revising its own guidance upward across multiple quarters rather than holding to initial-year projections. Tesla’s capex story has run on a different track, tied more to vehicle manufacturing capacity historically and only more recently reframed around AI compute for autonomy and robotics. The current divergence, one company raising guidance for the fourth or fifth consecutive cycle while another trims it, is the first time in this earnings season that the two spending philosophies have produced such a visible split in stock reaction on the same trading day.
Competitive Landscape: Divergent Spending Paths
The split between Tesla and Meta also reflects two different business models for monetizing AI hardware spend. Meta’s capex buys compute that directly serves billions of existing users through ad targeting, content recommendation, and now consumer AI products, giving the company a relatively short and visible path from spending to revenue. Tesla’s capex buys compute that serves a bet still being proven out: that autonomy and robotics will become large revenue lines on a timeline measured in years, not quarters.
That difference in monetization timeline likely explains why the market reacts so differently to similar-looking capex changes. Investors have years of data showing Meta’s infrastructure spending converts into ad revenue growth. Tesla’s AI infrastructure spending is still largely a forward promise, tied to the kind of robotaxi and robotics milestones discussed in Tesla Beats Estimates as Storage Hardware Offsets 3% Auto Drop. A cut to that unproven spending line carries more perceived risk than a cut to Meta’s proven one would, which is part of why a $130 billion-plus number moving up is cheered while a $20 billion number moving down is punished.
SpaceX, which is separately in advanced merger discussions with Tesla, adds another layer to how investors are reading Tesla’s hardware commitments this week. SpaceX’s own recent quarter showed strong growth tied partly to AI-adjacent compute demand, as covered in SpaceX’s $14.6B Quarter Fuels Starlink Mobile Pitch, and any combined entity would inherit a blended capex profile that markets have not yet fully priced.
What Analysts and Outlets Are Saying
The Financial Express framed the futures weakness as a direct function of Meta and Tesla earnings landing alongside broader AI-spending concerns, rather than any single macroeconomic trigger. That framing lines up with how narrowly the market moved: both major futures indexes fell by less than a full percentage point, consistent with a reaction to two specific earnings reports rather than a systemic shock.
Trading Economics data on the broader U.S. equity index backdrop shows the futures dip arriving in a market that had already been digesting a heavy run of earnings from AI-adjacent names this quarter. Investor’s Business Daily grouped the Tesla and Meta reports together with other names reporting the same week, underscoring how concentrated the market’s attention has become around a handful of capex-heavy reporters.
Market Impact: Why Capex Direction Now Moves Stocks More Than Earnings Beats
The most important pattern from this earnings cycle is that beating or missing on revenue and EPS is no longer sufficient to predict stock reaction for AI-exposed companies. Tesla beat on both lines relative to what the market had been bracing for, yet the stock fell because of a forward-looking spending number. Meta’s growth was strong, but it was the capex increase, not the 26% revenue growth alone, that appears to have reinforced investor confidence enough to push shares up more than 2% in extended trading.
This marks a shift in how the market prices AI infrastructure bets. For most of 2024 and into 2025, any capex increase was treated with some skepticism, as investors worried about overbuilding. By late 2026, the reaction has flipped for companies with a demonstrated revenue link to that spending, while companies whose AI payoff remains a future promise are now penalized for spending less, not more. That is a notable inversion, and it suggests the market has grown more comfortable underwriting large AI hardware budgets as long as there is visible revenue growth attached.
Predictions: Where AI Hardware Spending Goes From Here
- Expect more companies to break out capex guidance as a standalone, heavily scrutinized line item on earnings calls through 2027, following the template set by this week’s Tesla and Meta reports.
- Meta’s raised $135 billion-to-$145 billion range is likely to pressure component suppliers further, keeping DDR5 and high-end GPU pricing elevated into next year rather than easing as some builders had hoped.
- Tesla’s capex cut will likely draw continued questions on future calls about pacing for Optimus production and Full Self-Driving compute, especially given the company’s own prior framing of AI infrastructure as central to its valuation.
- Other major AI infrastructure spenders reporting later this earnings season will be watched primarily for capex direction, with revenue and EPS figures treated as secondary by comparison.
- If Tesla and SpaceX advance their reported merger discussions, a combined entity’s blended capex profile will become a new, closely tracked figure that markets have not yet had the chance to price in.
Risks and Uncertainties Ahead
Several open questions remain unresolved by the confirmed figures in these reports. It is not yet clear whether Tesla’s lower 2026 capex figure reflects a genuine pullback in ambition, a one-time timing shift of spending into later years, or efficiency gains that let the company do more with less hardware. None of those explanations has been confirmed, and reports should be treated as general until Tesla provides more detail on future calls.
Similarly, Meta’s raised guidance carries its own risk: a $135 billion-to-$145 billion commitment assumes continued revenue growth at or near the 26% pace reported this quarter. Any slowdown in ad revenue growth, consumer AI adoption, or broader digital advertising spending could leave Meta’s capex looking oversized relative to the business it is meant to support, a risk that applies to any company guiding aggressively higher on hardware spend during a period when AI monetization timelines remain unproven across the industry.
Frequently Asked Questions
Why did U.S. stock futures fall on October 22, 2026?
Nasdaq-100 futures fell 0.7% and S&P 500 futures fell 0.4% following Tesla and Meta Platforms earnings reports, with investors reacting to diverging capital-spending guidance from the two companies, according to Investor’s Business Daily.
What did Tesla report for the third quarter of 2026?
Tesla reported adjusted earnings of $0.48 per share on revenue of $28.05 billion, and guided 2026 capital expenditures to approximately $20 billion, a figure lower than prior spending expectations.
Why did Tesla shares fall after earnings despite beating expectations?
Tesla shares fell 3.3% in post-market trading specifically because the company cut its capital-spending forecast for 2026, not because of the headline earnings or revenue figures.
What did Meta Platforms report for the third quarter of 2026?
Meta reported earnings of $3.83 per share on revenue of $64.6 billion, up 26% year over year, and raised its 2026 capital-expenditure outlook to a range of $135 billion to $145 billion, up from a prior lower-end figure of $130 billion.
Why did Meta shares rise after raising its capex guidance?
Meta shares rose more than 2% in extended trading, a reaction tied to strong 26% revenue growth combined with a capex increase that investors read as confirmation of continued returns on AI infrastructure spending.
How does capital expenditure guidance affect AI hardware pricing?
Large capex commitments from companies like Meta translate directly into GPU, memory, and data center hardware demand, and have contributed to tight supply and elevated pricing for components such as DDR5 memory and high-end graphics cards throughout 2026.
Is Tesla’s capex cut a sign that AI spending overall is slowing down?
Not necessarily. Meta raised its own capex guidance in the same reporting window, suggesting the divergence reflects company-specific decisions rather than an industry-wide pullback in AI infrastructure spending.
What should investors watch next in this earnings cycle?
Analysts and outlets including Investor’s Business Daily and the Financial Express are watching capital-expenditure guidance from other major AI infrastructure spenders reporting later this season, treating that figure as a more significant signal than headline revenue or earnings-per-share results.
