Why Meta Is Struggling
Meta's recent stock weakness is best understood as a problem specific to Meta rather than a symptom of a broader malaise affecting all the big technology spenders. It is true that across mega-cap technology there is a massive amount of capital expenditure underway — likely on the order of roughly $750 billion this year, climbing to north of $800 billion in 2027 — and that this spending is weighing on free cash flow across the sector. But Meta faces a unique situation.
The central distinction is that Meta is not a hyperscaler, even though it is frequently lumped together with Google, Microsoft, and Amazon. Because it does not operate a cloud platform, it does not benefit from the "dollar-for-dollar spend equals revenue" dynamic that Nvidia's Jensen Huang frequently describes — the idea that infrastructure investment translates more or less directly into corresponding revenue. Instead, Meta has pursued a more dubious and expensive AI strategy. The company is spending hundreds of billions of dollars, and investors are simply less clear about what the eventual payoff will look like.
There are encouraging signs of return on investment within Meta's core digital advertising business. Beyond that, however, the effort to build large language models remains something Wall Street is genuinely uncertain about — and that uncertainty is sympathetic and reasonable, not merely skeptical.
The Vertical Integration Divide
Wall Street is currently sorting the major technology companies into two categories based on whether or not they are vertically integrated — what the market calls being "full stack." This framing goes directly to the heart of the concern about Meta.
Consider the companies that are praised for full-stack integration:
- Google is full stack all the way down — from Gemini, its model, through to the TPUs, the custom chips it designs and builds to run those models.
- Amazon has a comparable position with AWS: a whole host of layered-in AI tools sitting on top of its cloud, plus its Trainium chips. Notably, Amazon is now talking about selling those chips to third parties, putting it in direct competition with Nvidia.
This full-stack approach has earned more applause from Wall Street. Meta, by contrast, is approaching AI very differently. It is not so much building chips — though it likely would prefer to be doing more of that — but is instead trying to create the large language model itself, under names such as Muse Spark (previously referred to as Avocado). Mark Zuckerberg holds a vision that this work will improve the business over the long term, but that benefit has not yet shown up in the income statement.
It is worth stressing that AI is helping Meta's core digital advertising business. The advertisers who spend money on Meta's platforms — and there are over 100 million business customers spending across its family of apps — are seeing better return on investment. Crucially, however, those advertisers do not need large language models to obtain that improved ROI. As a result, the outcome of the much larger and more speculative model-building effort remains uncertain.
Can Meta Engineer a Turnaround Like Google's?
A key question is whether Meta can follow the path Google took. Google traded down sharply before experiencing a major narrative shift, and it subsequently became the top-performing member of the "Magnificent 7" heading into earnings season. Could Meta see a similar setup, or will its lack of vertical integration make a comparable rebound harder to achieve?
The answer offered here is firmly the former — Meta is expected to rerate higher. This is a high-conviction view: the position is one of continued ownership of Meta, which itself signals the bullish stance.
The Google precedent is instructive. About a year and a couple of months ago, Google's stock rerated all the way down to roughly 17 times earnings. The prevailing narrative at the time held that Gemini was not good enough in AI and that Google's core search business — the traditional "10 blue links" — was being disrupted by OpenAI, Claude, and similar tools. That narrative turned out to be false, and the decline proved to be a great buying opportunity. (For full disclosure, Google is also owned on behalf of clients.)
Meta is expected to undergo a similar narrative shift. This is unlikely to occur over the very immediate term — there are still issues that need to be worked out — but over the coming quarters and years, the case is compelling.
The Math Behind an $800 Price Target
The valuation case for Meta is described as straightforward. The stock has already rerated meaningfully: it sits about 30% off its highs, currently trading at around $565. At that level, it trades at roughly 17 times this year's earnings estimates and just over 15 times next year's estimates.
The path to a higher valuation rests on forward earnings. Projecting roughly $40 in earnings per share by 2028 and applying a deliberately conservative price-to-earnings multiple of 20 — conservative precisely because the stock trades at only 15 to 17 times today — produces a value of about $800 per share, achieved without much effort or heroic assumptions.
The underlying business supports this. The core business is growing at north of 20%. Meta is expected to generate around $260 billion in revenue this year and north of $300 billion next year, against a market capitalization now just under $1.5 trillion. Even measured on a price-to-sales basis, one does not need to make aggressive assumptions to arrive at meaningful upside.
Does the Spending Binge Concern or Comfort?
Asked directly whether Meta's enormous spending is comforting or concerning, the honest answer is neither — the spending does not provide comfort, but it is also not quite alarming.
The favorable view rests on several foundations of sound investing philosophy:
- A preference for owning companies with established market positions. Meta, alongside Google, is one of the largest players in digital advertising and has dominated that space for years.
- Digital advertising is a secular growth theme, providing strong structural tailwinds.
- Mark Zuckerberg is a founder-CEO, a setup that has historically tended to be a good one in technology — owning companies led by founders who are riding strong tailwinds.
The spending can be separated into two distinct buckets. A large portion of the investment is already visible in the form of boosted ROI for the companies spending on advertising across WhatsApp, Instagram, Facebook, and the rest of Meta's family of apps. That part is working. The big question mark lies in the spending devoted to building out the large language models.
The honest emotional posture toward that second bucket is a measured ambivalence: a little bit uncomfortable, because nothing has yet shown up to justify it and the case currently rests on a "trust me, trust me, trust me" message from Zuckerberg — but not deeply uncomfortable, because nobody understands the correct business direction for Meta better than Mark Zuckerberg himself. Betting alongside Zuckerberg on AI has been a good bet historically, and that bet is one worth continuing to make.
