The Thesis
Meta's core question is deceptively simple: does a 35 billion dollar capex program deliver a real moat or destroy returns? The market has already priced in some skepticism—we're at 24x forward earnings, not the 30x of 2021. My read is that Reality Labs loses money and always will, but that's irrelevant noise compared to what happens if AI-driven ad targeting cuts customer acquisition costs by even 15 percent. The advertising business growing 20 percent plus on a 60 billion dollar quarterly revenue base isn't something to dismiss with handwaving about capital intensity.
What changed for me on April 25 was the realization that even if Zuckerberg's infrastructure timeline compresses—even if they get meaningful monetization in 2027 instead of 2028—the risk reward flips. At 563.20 I sized the position at 60 percent of what I'd normally take. That's not conviction, that's respect for uncertainty.
The Setup
Q1 revenue expected to hit 43.5 billion. The beat will almost certainly be there; the miss narrative is on margins if capex guidance doesn't show a path to containment. But I've been through enough Meta earnings to know that Zuckerberg controls the message. If he frames 35 billion as temporary peak spend with a visible ceiling, the stock doesn't care that it's outrageous. The advertising business has always been about moat and scale. Throw in AI and you either get a widened moat or you get nothing. There's no "slightly cheaper CAC" outcome—it's binary enough.
The shorts expect Reality Labs to finally drag down consolidated margins. They will. The bulls expect advertising to offset it. I'm betting the bulls are right through 2027, with the caveat that if capex hits 40 billion without concrete Llama monetization, I'm out.
Entry and Execution
Bought 60 percent size at 563.20 on April 25. Straight equity, no leverage, no options. Stop is at 545. That's my line. If the earnings call signals capex creep without proof of concept on AI monetization, I'm taking the 3.3 percent loss and moving the capital to something with clearer ROI. I'm not emotional about it. The position is sized to let me sleep through volatility but not so large that I'm trapped.
I'll hold through the April 30 print. The expectation is that guidance will show modulation—maybe 38 billion instead of 35, or 35 with explicit milestones on AI revenue contribution. Neither is catastrophic if they're credible. If Zuckerberg suggests capex goes to 50 billion or higher, the risk reward inverts immediately.
Broader Context
The AI spending debate is the macro story nobody wants to admit is unresolved. Nvidia prints cash. The data centers get built. The foundational models improve. But nobody's actually made money at scale from this capex yet. OpenAI's still private. Anthropic's still private. Microsoft's seeing incremental search revenue from Copilot, nothing game-changing. So Meta spending 35 billion is either visionary or delusional. The market's hedging by pricing them at a discount, which is rational given the information asymmetry.
What I know is that Meta's advertising stack is the most sophisticated targeting engine on earth. The data is unmatched. If they can fuse that with proprietary AI models, the moat doesn't just widen—it becomes defensible. That's worth betting on at these prices, at this position size, with this stop.
What I'd Do Differently
I'd buy more if they guide capex down to 30 billion with a conviction argument on AI returns. I'd sell half on a 10 percent pop if it's capex guidance that sparks the rally instead of earnings beats. I'd cover if capex hits 40 billion without quarterly revenue from AI-driven products. Price target above 650 assumes advertising growth holds at 18 to 20 percent through 2027 and capex moderates to 25 billion by 2028.
If we break 545 on earnings volatility, I'll rebuy at 535 with a smaller tranche—that's where the risk reward justifies doubling down. Above 600 I'm taking profits; that's a 6.5 percent gain, which is fine preannouncement. The real money is in owning this through the next two quarters of earnings while the capex thesis either validates or evaporates.