Meta's Year of Efficiency was just the warm-up
The market still prices Meta like an ad company defending its turf. The real story is operating leverage — and an AI ad engine quietly compounding underneath it.
- → Flat headcount + low-20s revenue growth is driving margin expansion the Street still under-models.
- → The AI ad-loading ramp and Reels monetization gap closing faster than consensus.
- → $840 PT (+13%) on 22× '27 FCF — and the author thinks that's conservative.
Meta enters the second half of 2026 carrying a reputation it no longer deserves: the company that lights money on fire in the metaverse. That framing made sense in 2022. Today it is actively mispricing the stock.
Since I first published this thesis in August 2024, META is up 125% — and I think the easier money is still ahead. The multiple has re-rated, but the market has not yet absorbed what operating discipline plus an AI-native ad stack does to earnings power over the next eight quarters. Start with the chart that matters: not the metaverse losses everyone quotes, but the price action against the moment the call was made.
The operating leverage nobody modeled
Headcount is flat to down while revenue compounds in the low-20s. That is the entire story, and it is boring, and boring is exactly what compounds. Operating margin hit 41.5% in Q1 '2610-K , up from 38% a year earlier. Reality Labs is still a drag — but as a shrinking share of a growing pie.
The bears will tell you AI capex is the next Reality Labs — a bottomless pit dressed up as strategy. I disagree, and the difference is measurable: every dollar of inference spend today is attached to an ad auction that already clears at a positive ROI. That's not a moonshot. That's a flywheel.
"You don't have to believe in the metaverse to own Meta. You have to believe management has finally learned to count."
Which brings us to the part that actually drives the price target — the AI ad-loading ramp, the Reels monetization gap closing faster than consensus, and the buyback math that turns 22× into something closer to 17× on my numbers.
The model, the price target, and the bear case
Subscribe to read the full thesis, see every position, and get Sarah's calls the moment they publish — timestamped and tracked.
Disclosure. The author holds a long position in META, opened August 2024 and disclosed at publication. Conviction timestamps and locks every position the moment a thesis goes live. This is not investment advice.