Meta Platforms closed Wednesday at $777.59, capping its strongest month since July 2013 and leaving the company within touching distance of a $2 trillion valuation, and the entire move rests on a product that has not yet reported a dollar of revenue.
Here is what is actually established:
- Muse, Meta’s personal AI agent, launched in the United States on September 8. It sends email, books travel, fills out forms, and keeps working after the app is closed.
- It took the number one spot on Apple’s US App Store and recorded 2.8 million downloads in its first twelve days, per Apptopia data cited by Reuters as Wall Street repriced the company around the launch.
- Meta shares rose 36% across September, the best month since July 2013 on Bloomberg’s count, which put the company on the cusp of the $2 trillion club without crossing it.
- Meta guides to $130 billion to $145 billion of capital expenditure in 2026.
- Meta has not disclosed Muse revenue, Muse pricing, or a Muse subscriber count. There is no paid tier in market.
Every outlet covering this rally reported the download number and the price-target raises that followed it. What none of them did was put the banks’ own bull case next to the company’s own spending line. We did, and the arithmetic is unkind: the most optimistic published revenue scenario for Muse, two years out, covers roughly seven percent of a single year’s capital budget.
The Bull Case, Stated in Full
Jefferies raised its Meta price target to $875 on the strength of a specific, and specifically enormous, chain of assumptions. The bank modelled Muse reaching one billion users by the end of 2027, with at least three percent of them converting to a $30 monthly subscription. That produces 30 million paying subscribers and $10.8 billion in annualised revenue.
Take that scenario entirely at face value. Assume the billion users arrive, assume the conversion holds, assume the $30 price survives contact with a market where OpenAI, Google and Anthropic are all selling comparable assistants. The result is $10.8 billion of annual revenue against capital expenditure that Meta itself has already guided to $130 billion to $145 billion for 2026 alone.
| Line | Figure | Source |
|---|---|---|
| Muse downloads, first 12 days | 2.8 million | Apptopia, via Reuters |
| Users required by end-2027, Jefferies bull case | 1 billion | Jefferies |
| Implied growth multiple | about 357x | BTN calculation |
| Bull-case annualised revenue, 2027 | $10.8 billion | Jefferies |
| Meta capex guidance, 2026 alone | $130bn to $145bn | Meta guidance |
| Bull case as share of one year’s capex | 7% to 8% | BTN calculation |
| Q2 free cash flow | $784 million, down from $8.55bn | Meta Q2 |
That 357x figure is the one nobody is printing. Muse needs to multiply its user base by more than three hundred and fifty times in roughly fifteen months for the optimistic case to land. Not the base case. The optimistic one.
What the Cash Flow Statement Already Said
The spending is not a forecast. It is happening now, and it is already visible in the accounts. Meta’s second-quarter capital expenditure hit $30.12 billion, up 82% year over year. Operating margin fell to 31% from 43%. Free cash flow collapsed to $784 million from $8.55 billion in the same quarter a year earlier, a decline of more than 90%.
Investors understood this perfectly well four months ago. When Meta first put the $145 billion capex figure in front of the market, the stock fell 10% and JPMorgan downgraded it on precisely the return-on-investment question that has now been declared answered. Nothing about the spending has changed since. What changed is that a consumer app went to number one on an app store.
This is also not Meta’s first paid AI subscription. The company rolled paid AI tiers across Instagram, Facebook and WhatsApp in May while cutting 8,000 jobs. That rollout did not reprice the stock by half a trillion dollars, because a download chart is easier to point at than a monetisation curve.
The Hardware Tell
At Meta Connect on September 23, Zuckerberg unveiled the Muse Charm, a keychain-sized pendant with a roughly two-inch touchscreen and a fingerprint sensor, built to run Muse without a phone. He also showed $1,299 VR glasses, a third-generation Ray-Ban Meta, and audio-only smart glasses.
Read the Charm announcement closely and it tells you where Meta itself thinks the business is. The device ships in December. Meta has not set a price. People familiar with the plans have indicated it will land somewhere in smartwatch territory, which is a range, not a number.
A company that had modelled the subscription economics of its flagship AI product would know what the companion hardware costs. Shipping a holiday device with an undecided price three months out is what it looks like when the product roadmap is running ahead of the revenue model. Zuckerberg told the keynote audience that Muse would “grow into the personal superintelligence that billions of people are going to use.” Billions is the vision. One billion by the end of 2027 is the sell-side model. Neither is a price list.
Who Is Served by the Framing
The framing that took hold in the last fortnight is that Muse validates the capex. It is worth asking who that framing serves.
It serves the analysts, who spent a year fielding the return-on-investment question and can now point at an App Store rank instead of a revenue line. It serves Meta, whose spending is now read as vision rather than risk. It serves every holder who bought above $570 in August.
It does not serve anyone trying to work out what this company earns. Free cash flow fell more than 90% year over year in the most recent quarter. That is the measured fact. The $10.8 billion is a scenario, and it belongs to 2027.
Here is where BTN stands. The Muse rally is not irrational, but it is not a subscription business being repriced. It is an option being repriced, and options are worth something. The error is in the translation. Wall Street has added something close to half a trillion dollars of market value in a month on evidence that consists of a download count and a chart position, while the only published revenue scenario that would justify any of it concedes it cannot cover a tenth of one year’s spending.
The responsibility sits with the sell side. Wells Fargo lifted its target to $796 from $640 as the download numbers landed. Raising a price target by $156 on an App Store rank, then publishing a model whose own output is dwarfed by the cost line it is meant to justify, is not analysis. It is narrative with a spreadsheet attached.
What Would Actually Settle It
Meta can end the argument whenever it likes. Three disclosures would do it: a Muse price, a paid-conversion rate, and a retention curve past ninety days. Downloads are not users. Users are not subscribers.
Until those numbers exist, the $2 trillion line is a bet that a fifteen-month, 357x growth run is under way. Meta’s third-quarter results are the first real test.
Watch the cash flow statement, not the app chart.