Meta Platforms conservative fair value ~$810/share (feedback welcome)

Would appreciate pushback on the assumptions below. I'm currently landing at fair value of \~$810/share vs. \~$640/share. For context, META is one of my largest positions in my portfolio (\~$3mm cost, check my post history for context)

Link to my model on Google Sheets: [https://docs.google.com/spreadsheets/d/1dlt-9YaB1uV3RF4jz6snuS2GCahCeZPJHuCKcCUIm10/edit?gid=1127915579#gid=1127915579](https://docs.google.com/spreadsheets/d/1dlt-9YaB1uV3RF4jz6snuS2GCahCeZPJHuCKcCUIm10/edit?gid=1127915579#gid=1127915579)

* Assuming revenue grows 20% annually (20% FoA growth and Reality Labs stays flat) through 2030 with \~100bps of EBITDA margin expansion. This is probably conservative as I'm essentially assuming 100% of Opex is fixed and growing in-line with revenue when, in reality, you'd probably get some operating leverage on the cost structure.
* Capex is obviously the biggest variable - for 2026 and 2027, I'm assuming the midpoint of management's latest guidance for 2026 ($135bn) and street estimates for 2027 ($200bn). The biggest driver of the valuation is what I should assume in 2028. Currently I’m assuming it steps down to $150bn and then grows 5% annually terminally. However, this may be overly punitive as I'm assuming no incremental cash flows from potential compute leasing revenue or acceleration in advertising growth.
* Model is highly sensitive to what 2028 capex is - in the most bottom sensitivity table, if 2028 capex steps down to $100bn, then that implies share price north of $1,100/share
* Cash flows discounted at 10% WACC
* Using a multi-stage growth where I assume cash flows past 2030 grow at 10% over 10 years and then steps down to GDP-type growth (\~2%) afterwards