What Meta's $50B AI push means for the meta stock price

Meta is pouring about $50 billion into Hyperion, a U.S. AI data center meant to power models, agents, and surplus compute—while full-year capex could hit $130–$145 billion. That spend has weighed on the meta stock price near the mid-$500s, even as ads grow and analysts still see upside toward $750–$860.

Investors keep asking the same question: when does all this AI infrastructure start paying for itself? Recent reporting from Simply Wall St, Barchart, and GuruFocus sketches the answer—and why the selloff may matter for long-term holders looking at wealth hacks and passive income themes tied to mega-cap tech.

Key Takeaways

• Meta is advancing Hyperion, a planned $50 billion U.S. data center aimed at AI workloads, with about 5 GW of capacity and in-house chips in the broader build-out story.
• Full-year 2026 capex guidance sits at $130–$145 billion; free cash flow fell more than 90% last quarter even as revenue rose 28%.
• Management told BNP Paribas selling “intelligence” (models, agents, products) is more profitable than renting compute, which is a backup plan.
• The meta stock price has lagged—down about 25% over the past year—while Wall Street’s mean target is about $750 and some firms see $855–$860.
• GuruFocus pegs META as roughly 33.6% undervalued versus its GF Value of $842.41, though insider selling remains a caveat.

What Is Meta Actually Building With Its $50 Billion AI Push?

According to Yahoo Finance’s Simply Wall St coverage, Meta Platforms is moving ahead with Hyperion, a planned $50 billion data center project in the United States. The facility is intended to rank among the country’s largest, focused on Meta’s AI workloads and operational flexibility.

That infrastructure push sits alongside talent moves. Meta has recruited AI researcher Luke Metz into its Superintelligence Labs unit, adding senior capacity to advanced research. Simply Wall St frames Hyperion and the hire as reinforcing a familiar narrative: heavy AI infrastructure and talent spending come first, with AI subscriptions and enterprise agents expected to monetize that compute over time.

The same analysis notes Meta is building toward about 5 GW of capacity and developing in-house chips. Longer term, the company could sell access to its models and surplus compute on top of advertising. The practical test, Simply Wall St argues, is whether Meta starts disclosing hard AI-related revenue or usage figures in coming earnings reports.

Why Has the Meta Stock Price Fallen Despite Strong Ads?

Barchart reports Meta plans to invest between $130 billion and $145 billion this year. Free cash flow collapsed, falling more than 90% last quarter. Revenue is still growing fast—up 28% in the June quarter—but investors are uneasy about a company pouring nearly every dollar it earns back into data centers.

Second-quarter fiscal 2026 results, reported July 29, showed revenue of $60.8 billion, above a $60.2 billion forecast. Family of Apps revenue was $60.4 billion, ad revenue $59.4 billion, Reality Labs $431 million, and other revenue $1 billion. Diluted EPS was $6.18, down 13% year over year and below the $7.17 consensus.

Total expenses hit $42 billion, up 55% year over year. Capital expenditures were $31.1 billion, including principal payments on finance leases. Meta guided third-quarter revenue to $61–$64 billion and raised the lower end of full-year expense outlook to $165–$169 billion, citing a $2.4 billion legal charge. Capex guidance narrowed to $130–$145 billion from a prior $125–$145 billion range.

Over the past year, META has declined almost 25% while the S&P 500 gained roughly 19%. Year to date, META is down about 15% versus an index gain near 12%. Barchart notes the forward P/E of about 19 times sits below Meta’s five-year average of 22 times, while the price-to-sales ratio of 6.9 times is roughly in line with its five-year average.

How High Could the Meta Stock Price Go From Here?

After meeting Meta executives, BNP Paribas maintained an Outperform rating with an $855 price target, versus a share price then around $566, Barchart reported. Management told the firm selling “intelligence”—AI models, agents, and finished products—is far more profitable than renting out compute. Renting spare capacity is described as a backup if Meta’s own needs run lower than expected.

Barchart’s author argues that if the forward P/E simply returned to its historical average, META could reach roughly $700–$750 without needing upside surprises. Clear monetization of AI spending could expand the multiple further. Analysts expect a modest 4% EPS drop in fiscal 2026 as spending squeezes margins, then about 21% earnings growth to $34.47 per share in fiscal 2027. Meta holds roughly $90 billion in cash against about $84 billion in debt.

Other Street views line up bullishly. Citi’s Ronald Josey reiterated a Buy with an $850 target. Bernstein’s Mark Shmulik kept Outperform at $800. Across 55 analysts, the consensus is Strong Buy, with a mean target of $750.37 (about 33% upside) and a high of $1,000 (about 77%).

Separately, GuruFocus reported Evercore ISI’s Mark Mahaney maintained Outperform and raised his price target from $820 to $860. At a quoted price of $559.02, GuruFocus’ GF Value stood at $842.41—about 33.6% undervalued—with a GF Score of 92/100. Meta’s market cap was about $1.42 trillion, serving nearly 4 billion monthly active users. A P/E of 21.06 versus a five-year median of 26.07 supports the discount thesis, though $24.54 million in insider sells over three months is a caution flag.

What Should Investors Watch Next?

The near-term story is execution risk versus monetization. Capex and operating costs can run ahead of direct AI revenue for years. Simply Wall St stresses that quantified disclosure on AI subscriptions, enterprise agents, or external infrastructure access would strengthen the compute-monetization case.

Until then, the meta stock price largely reflects a tug-of-war: a still-growing ad engine funding a historic AI build-out, and a market that wants proof the $50 billion Hyperion bet—and the much larger annual spend—will compound earnings, not just capacity.

This article is for informational purposes only and is not investment advice. Always do your own research before buying or selling any security.