Meta Stock Under Pressure: 4 Signals Investors Can’t Ignore as AI Spending Takes Center Stage

Meta stock is sending an unusual mixed message: a business that still posts strong growth and margins, yet a share price that has cooled sharply as investors debate the size, timing, and discipline of its AI spending. The tension is not only about near-term market sentiment, but about whether a new investment cycle can deliver durable returns after a costly metaverse chapter. With Meta down about 24% from its 52-week high and negative year-to-date, the market is forcing a reset in expectations.

Why Meta stock is sliding even as the core business looks strong

The immediate paradox is clear in the numbers that have been put on the table. Meta generated 22% revenue growth last year and posted an impressive 30% profit margin, even while funding major long-horizon projects. Yet Meta stock has struggled recently, underperforming the broader market and sliding about 24% from its 52-week high of $796. 25. The shares are also down about 9% this year, reflecting a shift in investor appetite for big technology names amid broader bearish sentiment toward the sector.

On valuation, the picture is similarly split. Meta trades around 25 times trailing earnings, roughly in line with the S& P 500 average multiple of 24, which supports the argument that the stock is not obviously overpriced on a headline basis. Another valuation framing pegs a last close share price of $593. 66 against a “narrative fair value” estimate of $723. 11, presenting Meta as undervalued in that model. At the same time, the earnings-based lens complicates the story: a P/E of 24. 8x sits above the US Interactive Media and Services industry average of 14. 7x, yet below a peer average of 29. 4x.

The market’s message appears less about what Meta is today and more about how credibly it can convert a new wave of AI investment into results without repeating past spending missteps.

Meta Stock and the new AI capex era: scale, timing, and the “proof gap”

The dominant driver of the debate is capital intensity. Meta has laid out extremely large spending ranges tied to AI infrastructure and ambition. One set of figures signals planned capital expenditures of $115 billion to $135 billion for 2026, as the company pursues “superintelligence, ” compared with $72 billion of total capital spend last year. Separately, Meta has also flagged a $60 billion to $65 billion capex commitment for 2025, described as roughly twice its 2024 capex of $38 billion.

What investors are wrestling with is not whether AI is strategically important, but the “proof gap” between spending now and demonstrating returns later. Data center construction can take 2–3 years from site selection to operational capacity, meaning 2025 capex aimed at construction may primarily add capacity in 2027–2028. That lag creates a cash flow timing challenge even for a company with significant financial strength.

Meta’s AI infrastructure build is described as heavily focused on data center construction, AI training and inference hardware, and network infrastructure. The company has been building large-scale GPU clusters for AI model training, including a cluster with over 100, 000 H100 GPUs that was announced in 2024. It is also developing custom silicon through its MTIA (Meta Training and Inference Accelerator) chip program, designed for recommendation and advertising inference tasks, with the stated goal of improving cost efficiency and reducing dependence on third-party GPUs for certain workloads.

These efforts are coherent as an industrial strategy, but they also raise a straightforward investor question: how soon will the economics show up in measurable performance and cash generation? Without that evidence, Meta stock can remain vulnerable to concerns that spending is racing ahead of returns.

Reality Labs, governance risk, and valuation: the longer shadows on Meta stock

AI spending is not being evaluated in isolation. Meta’s history of heavy investment in the metaverse remains an active reference point for skeptical investors. Last year, Reality Labs incurred losses totaling more than $19 billion. While the company has been described as backing away from metaverse efforts as it pivots more toward AI, that earlier period reinforces the perception that capital allocation discipline is a key swing factor.

At the same time, the company’s evolution into “global digital infrastructure” changes how it is judged. One narrative view argues that growth alone is no longer sufficient to explain valuation; durability, governance, and legal exposure play a larger role in determining long-term value. In that framework, a fair value estimate of $723. 11 is tied not just to scale but also to a “liability profile, ” implicitly acknowledging that the platform’s reach brings broader exposure to regulatory or legal pressure.

Importantly, these are not just abstract risks. The same valuation narrative explicitly flags heavier regulatory or legal pressure around Meta’s global reach as a key downside variable, alongside weaker returns from Reality Labs’ reported US$2. 2 billion revenue contribution. That combination—big reinvestment plus long-tail governance risk—helps explain why a profitable company can still trade with caution embedded in the price.

The market is, in effect, asking Meta to do two things at once: preserve advertising dominance while proving that high AI capex will be managed with a sharper internal hurdle rate than the metaverse era.

Expert perspectives: how to read the next chapter

Meta’s investment posture is repeatedly tied to CEO Mark Zuckerberg’s conviction that AI will be transformative for the company’s advertising business and that the window to invest at scale is now. The company’s expense framework also reflects the breadth of its ambitions across AI, infrastructure, and long-horizon bets that may not pay off for years.

Within Meta, AI research is described as substantial, including work through the company’s FAIR (Fundamental AI Research) lab, which has produced the LLaMA open-source language model series. The strategy of open-sourcing AI models is framed as an ecosystem play, positioning Meta to focus on proprietary implementation advantages in areas such as recommendations and advertising systems.

For investors, the practical read-through is that the next decisive inputs for Meta stock will be evidence that the AI buildout strengthens the advertising engine and that the cost curve improves as infrastructure scales. The company’s large workforce—described at approximately 70, 000–75, 000 people globally, with engineering and product talent as the majority—underscores that this is a multi-year execution story, not a single-quarter narrative.

Factually, Meta’s profitability and revenue momentum show the core business has remained resilient. Analytically, the stock’s weakness signals that resilience is no longer enough: the market wants proof that the new spend cycle converts into durable advantage without recreating the metaverse-style drag.

What happens next: a wait-and-see test for Meta stock

The near-term direction for Meta stock hinges on whether Meta can show its aggressive AI push is paying off, particularly as concerns around high AI spending persist and tech sentiment remains fragile. With shares down about 9% this year and momentum cooling after a strong multi-year run, the company’s next phase looks less like a growth-at-any-cost story and more like an execution-and-credibility test.

If the market is applying a “quality premium” or building in “future risk” remains an open question. But the next challenge is clear: can Meta translate unprecedented AI capital commitments into visible return metrics fast enough to rebuild confidence, or will Meta stock continue to reflect a skepticism shaped by the last era of expensive ambition?