Morgan Stanley just cut its price target on Meta Platforms. Not by much. From $660 to $650. But the signal matters more than the magnitude, because it reflects a growing tension on Wall Street between faith in Meta’s artificial intelligence ambitions and hard questions about when those ambitions will actually produce returns.
The trim, reported by Yahoo Finance, came as analyst Brian Nowak maintained his overweight rating on the stock while acknowledging that Meta’s capital expenditure plans have ballooned to a point that demands scrutiny. Meta now expects to spend between $60 billion and $65 billion on capital expenditures in 2025 alone — a figure that has climbed repeatedly over the past year and one that dwarfs the spending profiles of most companies in any industry, let alone social media.
This isn’t a downgrade. Morgan Stanley still believes in the stock. But the math is getting harder to ignore.
Meta shares have had a turbulent stretch. After surging through much of 2024 on enthusiasm around AI integration across its family of apps — Facebook, Instagram, WhatsApp, Messenger — the stock pulled back sharply in early 2025 amid broader market volatility and rising anxiety about whether Big Tech’s AI spending spree would deliver commensurate revenue growth. Shares traded around $595 in recent sessions, well below the highs north of $700 reached earlier this year but still up significantly from 2023 lows.
The central question facing Meta investors is deceptively simple: Is $65 billion a year in capex justified?
Mark Zuckerberg thinks so. On Meta’s most recent earnings call, the CEO was characteristically direct about his conviction that AI will transform every layer of the company’s business — from ad targeting and content recommendation to the long-term development of the metaverse and augmented reality hardware. He described 2025 as a “defining year” for AI infrastructure investment, framing the spending not as discretionary but as existential. Companies that don’t invest aggressively now, he argued, will find themselves permanently behind.
Wall Street has heard this before. The last time Zuckerberg made a massive, conviction-driven capital allocation bet — pouring tens of billions into Reality Labs and the metaverse starting in 2021 — investors revolted. Meta’s stock lost more than 75% of its value between late 2021 and late 2022. The company eventually course-corrected with its “Year of Efficiency” in 2023, slashing headcount, cutting projects, and refocusing on profitability. The stock recovered spectacularly.
Now the spending is back. Different label, same scale.
Morgan Stanley’s Nowak, in trimming his target, pointed to the widening gap between Meta’s capital intensity and its near-term revenue trajectory. While Meta’s advertising business remains extraordinarily profitable — the company generated over $160 billion in revenue last year with operating margins above 30% — the incremental dollars being poured into AI data centers, custom chips, and GPU clusters are growing faster than the incremental revenue those investments have so far produced. The payoff is theoretical. The costs are very real.
And Meta is far from alone in this predicament. Alphabet, Amazon, and Microsoft have all dramatically increased their AI-related capital spending, collectively committing hundreds of billions of dollars over the next several years. But Meta’s position is arguably the most precarious of the group because it lacks the diversified revenue streams of its peers. Alphabet has Google Cloud. Amazon has AWS. Microsoft has Azure. Meta has advertising. Period. If the AI spending doesn’t translate into meaningfully better ad products — or open up entirely new revenue lines — the return on investment calculus could turn ugly.
Recent reporting from Reuters and Bloomberg has highlighted growing skepticism among some institutional investors about the timeline for AI monetization across Big Tech. Several large fund managers have reportedly trimmed positions in the so-called Magnificent Seven stocks, citing valuation concerns and the risk that AI spending could compress margins for longer than consensus expects. Meta has been a particular focus of this rotation, given its concentration in a single revenue stream.
To be fair, Meta has shown early evidence that AI is already improving its core business. The company’s AI-driven recommendation algorithms have meaningfully increased time spent on Instagram Reels and Facebook’s main feed, which in turn has boosted ad impressions and pricing power. Meta’s Advantage+ automated ad products, which use machine learning to optimize campaign targeting and creative, have seen rapid adoption among advertisers. And Meta AI, the company’s consumer-facing chatbot built on its Llama large language model, has reached hundreds of millions of monthly active users faster than almost any product in Meta’s history.
These are real results. But they’re also incremental improvements to an existing business, not the kind of step-function revenue growth that would justify a $65 billion annual spend.
The bull case rests on what comes next. Meta’s leadership has outlined a vision where AI agents — autonomous software programs capable of performing tasks on behalf of users and businesses — become a primary interface for commerce, customer service, and content creation across its platforms. If Meta can position itself as the infrastructure layer for AI-powered business interactions at the scale of its 3.3 billion daily active users, the revenue opportunity would dwarf current advertising economics. That’s the bet.
But it’s a bet, not a certainty. And the market is increasingly pricing it as such.
Morgan Stanley’s revised $650 target implies roughly 9% upside from recent levels — a decent return, but hardly the kind of aggressive call that characterized Wall Street’s Meta coverage during the AI euphoria of mid-2024. Other major banks have also moderated their enthusiasm. JPMorgan recently reiterated its overweight rating but flagged capex risk as a key variable. Bank of America maintained its buy rating while noting that investor sentiment has shifted from “AI optimism” to “AI accountability” — meaning the market now wants to see proof that these investments are generating measurable returns.
The broader context matters here too. President Trump’s tariff policies have injected fresh uncertainty into the global advertising market, which is Meta’s lifeblood. While Meta’s ad business has proven remarkably resilient to macroeconomic headwinds in recent years — partly because of the shift from traditional to digital advertising, and partly because of the AI-driven improvements to ad performance — a sustained trade war or economic slowdown could pressure ad budgets in ways that make $65 billion in annual capex feel less sustainable.
There’s also the competitive dimension. OpenAI, backed by Microsoft, continues to push the frontier of large language models with GPT-5 and beyond. Google’s Gemini models are improving rapidly. Apple is integrating AI features across its device ecosystem in ways that could reduce user engagement with Meta’s apps. And a wave of AI startups — from Perplexity to Anthropic to Mistral — are competing for the same talent and compute resources that Meta is spending billions to secure.
Meta’s open-source strategy with Llama gives it certain advantages in developer adoption and community goodwill. But open-source also means competitors can build on Meta’s work without paying for it, which complicates the monetization picture.
So where does this leave investors?
The consensus view on Wall Street remains bullish, but the conviction has softened. Of the 60-plus analysts covering Meta, the vast majority still rate it a buy or equivalent. The average price target sits around $670, according to data compiled by Bloomberg — above Morgan Stanley’s revised figure but well below the most optimistic targets north of $800. The stock trades at roughly 22 times forward earnings, a premium to the S&P 500 but a discount to its own five-year average multiple.
That valuation reflects a market that’s cautiously optimistic but no longer willing to give Meta a blank check. Investors want to see the 2025 capex translate into concrete, measurable improvements in revenue growth and margin expansion. They want evidence that Llama is becoming a platform, not just a research project. They want proof that AI agents can generate real commercial value at scale.
Meta’s next earnings report will be a critical test. The company is expected to provide updated guidance on both revenue trajectory and capital spending plans. Any hint that the $60-65 billion capex range could move higher — or that revenue growth is decelerating — would likely trigger another round of target cuts and potentially a more meaningful selloff.
Conversely, if Meta can demonstrate that its AI investments are accelerating top-line growth and opening new revenue streams, the stock has significant room to re-rate higher. The company’s underlying business remains a cash-generating machine. Free cash flow exceeded $40 billion last year even after massive capital expenditures. That’s a financial position most companies would envy.
Morgan Stanley’s $10 trim isn’t a vote of no confidence. It’s a recalibration. A recognition that the gap between AI ambition and AI reality is still wide, and that the market’s patience — while not exhausted — is no longer infinite. Meta has the resources, the user base, and the technical talent to make its AI bet pay off. Whether it will, and how quickly, remains the most consequential open question in Big Tech.
The answer is worth roughly $1.5 trillion. That’s Meta’s current market capitalization. And every dollar of that valuation now depends, to a degree unprecedented in the company’s history, on whether artificial intelligence can do what Mark Zuckerberg says it can.
No pressure.
WebProNews is an iEntry Publication