Discerning Signal from Noise: Why Meta's Advertising Franchise is Materially Mispriced

The mispricing is driven by obfuscating consolidated reporting, increasing capex, and fears of capital incineration. These fears are dispelled by the facts, with ~39% upside to base case IV ($795).

Introduction

The Variant View’s north star for analysis is: focus on what’s knowable and important. That is a particularly useful rule, especially when it comes to $META; a polarising business that seemingly becomes the subject of intense fear from market participants every two to three years.

As we all know, the last cycle of fear occurred in 2021: the catalyst then was Zuckerberg’s relentless capital incineration in pursuit of the Metaverse. Discontent with that particular capital allocation decision was warranted and has been vindicated 5 years later; not much has come of the tens of billions of dollars spent on that project. What’s more, existing shareholders at the time then suffered a torturous ~70% decline in Meta’s share price as their fellow business owners succumbed to fear and leapt from what they perceived to be a sinking ship. The sinking ship instead turned out to be a leviathan rising from the deep, with a trough to peak recovery of ~700% over the next 3 years.

During that torturous ~70% decline, what could have possibly grounded existing shareholders and incremental purchasers of Meta’s stock to cling to the rails of what others vehemently claimed to be a sinking ship? The north star at that time — the fact that was knowable and important above all else — was the solidity of Meta’s advertising business. The economic characteristics of that business — particularly its two-sided network effects and near-zero marginal costs of distribution — were completely unaffected by the Metaverse spending. That was the insight that grounded me and prompted me to purchase $Meta’s stock at an average price of $158 over that period (though I wasn’t spared from torture — the stock promptly went to ~$80).

History doesn’t repeat but it does rhyme. The spectre of capital incineration has again returned to instill fear in market participants. This time, the primary concern is that the FY26 capex guide of $115B-$135B is a Metaverse 2.0 spending spree. The enormity of the capital incinerated in 2021 appears to be exerting a kind of availability bias in the market — many participants are genuinely convinced that Meta’s management, and Zuckerberg in particular, are not good capital allocators. They’re like Twain’s cat: having sat on a hot stove once, they’re not willing to sit on any stove, whether hot or cold, ever again. To carry on the metaphor, the stove isn’t hot this time: Meta’s AI infrastructure capex spending is demonstrably producing improvements in its advertising segment. Furthermore, an analysis of Meta’s return on incremental capital expenditures over the past 5 years does not support the conclusion that management are poor capital allocators.

So, if we’re in broadly similar circumstances to 2021, at least in terms of fear and mispricing (there are some different issues at play and the mispricing is not as severe), what’s knowable and important this time around? Once again, it’s the advertising segment’s economics, which are stronger than ever: the economic moat is deepening, profitability is accelerating, but its operating results are obfuscated by consolidated financial reporting, which includes operating income losses generated by the Reality Labs segment. There is one caveat, however: the game has changed. The primary long-term risk to Meta’s business is whether agentic AI disrupts the advertising discovery model before Meta can position itself as the agentic advertising layer.

Overall, my thesis is that Meta is once again materially mispriced. The mispricing is broadly being driven by obfuscation of the advertising segment’s operating results; fear that FY26 capex is the Metaverse 2.0; and the belief that Meta’s management are poor capital allocators.

I estimate Meta’s intrinsic value per share to be ~$795 (+39% from last close price of $572.13) and I view fears about unproductive FY26 capex spending as unsubstantiated by my calculation of management’s FY2021-FY2025 return on incremental capex track record.

My full analysis is available as a downloadable PDF document for paid subscribers of the blog beyond the paywall. Topics of analysis covered in the document include:

1. How Meta's consolidated financials obscure the profitability of its advertising segment; the implications of this for Meta's current valuation; and a brief description of the increasing strength and deepening economic moat of the advertising flywheel;

2. Why 2026 capex is not the Metaverse 2.0, including:

   (a) evidence of revenues already being generated by $Meta's AI capex;
   (b) what this capex is for and (to the best that the financials allow) how it's being directed as between different projects within the company;
   (c) management's falsifiable commitments to prospective returns on increasing capex for FY26;
   (d) analysis of Meta's returns on incremental capex track record over the past 5 years;
   (d) brief assessment of capital returns over the past 5 years;

3. Key risks adjudicated against evidence, including the agentic AI structural threat and litigation risk, among others;

4. Brief description of unpriced catalysts;

5. The market's current mispricing of the business, the underlying assumptions that the mispricing implies, and how those assumptions are contradicted by the evidence;

6. Valuation analysis, including bear ($571, -0.2%), base ($795, +39%), and upside ($1,015, +77%) cases relative to the last close price of $572.13.

To those who decide to read, thank you for your support. I hope the report is insightful and useful to you. I put a lot of time and energy into it. Constructive feedback is always welcome.