Meta (META): Free Cash Flow Fell 91% While Revenue Grew 28% — Moat or Money Pit?

Meta Platforms, Inc. (META) Q2 2026 — Meta Platforms (META) reported Q2 2026 (quarter ended June 30, 2026) after the close on July 29: revenue $60.80B, +28% YoY (+27% constant currency), a beat — but diluted EPS fell 13% to $6.18, missing the Street by roughly a dollar, and operating margin collapsed from 43% to 31%. Total costs and expenses rose 55% to $42.03B, including $2.40B of legal charges and $1.18B of severance from the May headcount reduction; R&D alone rose 67% to $21.66B (35.6% of revenue). The number that moved the stock: capex was $31.08B in a single quarter (51% of revenue), leaving free cash flow of just $784M, down 91% from $8.55B. Full-year 2026 capex guidance narrowed to $130-145B (vs $72B spent in 2025), implying $79-94B in H2 alone. Meta bought back ZERO stock in Q2 and zero in H1 (vs $22.9B in H1 2025) and instead issued $24.9B of long-term debt, taking total debt to $83.7B against $90.3B of cash and securities — net cash of only ~$6.6B, down from ~$23B six months ago. Stock comp was $7.66B, +58%, while headcount FELL 1% to 75,472 (~$406k of stock per employee annualized vs ~$254k a year ago). Under-covered angles: US & Canada is 44% of revenue and grew 31.6% with price per ad +20% on only +9% impressions (genuine pricing power), while Asia-Pacific price per ad grew just 1% and Europe just 10% (DMA pay-or-consent drag); Reality Labs lost $4.62B on $431M of revenue, a wider loss than last year and roughly $97B of cumulative operating losses since 2019; and D&A of $6.36B (a ~$25B annual run-rate) must eventually climb toward the $137B annual capex rate — if depreciation were already there, this quarter's $18.8B of operating profit would be about $2B. Our owner-earnings DCF (maintenance capex charged at $45B/yr, stock comp expensed, 10.5% discount rate) lands a probability-weighted fair value near $610 vs ~$527.34 in the July 30 pre-market (the pre-print close of $585.61 is the wrong basis — this printed AMC). Our call: BUY, 3/5. Wall Street's aggregate target is $724.50 (Buy, 65 analysts), but every analyst who updated after the print CUT: JPMorgan $640 (Neutral), UBS $715, Goldman $725, TD Cowen $750, Citi $800, BofA $810 — six cuts, zero raises, median ~$740. We AGREE on direction and DIFFER on value: our $610 sits below even the lowest fresh target. Meta Platforms just printed one of the strangest quarters in mega-cap history: revenue accelerated to $60.80 billion, up 28% year over year — and free cash flow collapsed 91% to $784 million. Q2 2026 (ended June 30, 2026) delivered a revenue beat and a profit miss in the same release. Operating income fell 8% to $18.78 billion, diluted EPS fell 13% to $6.18, and operating margin dropped from 43% to 31%. Strip out the $2.40B of legal charges and $1.18B of severance and margin is still only ~36.8%, because R&D rose 67% to $21.66 billion — 35.6 cents of every revenue dollar. The line almost nobody leads with: Family of Apps operating income FELL 6% (to $23.39B) on revenue that grew 28%; the core ad franchise earned fewer absolute dollars. The cause is capital spending. Capex was $31.08 billion in one quarter — 51% of revenue — and full-year guidance narrowed to $130-145 billion, implying $79-94 billion in the second half alone, nearly double the first. Meanwhile depreciation and amortization is only $6.36 billion a quarter, a ~$25 billion annual run-rate against a ~$137 billion annual spend rate. That gap is the depreciation cliff: if depreciation were already where the spend rate implies, this quarter's $18.8B of operating profit would be roughly $2B. Three under-covered angles decide this stock. First, pricing power is real but concentrated: in the US & Canada — 44% of all revenue — price per ad rose 20% on just 9% more impressions, while Asia-Pacific price per ad rose 1% and Europe only 10% (the DMA pay-or-consent drag). Second, capital allocation reversed: Meta repurchased zero stock in Q2 and zero in H1 (vs $22.9B a year earlier) and issued $24.9 billion of long-term debt instead; net cash is down to ~$6.6 billion from ~$23 billion, and stock comp of $7.66 billion (+58%) on a headcount that FELL 1% works out to roughly $406,000 of stock per employee per year. Third, Reality Labs lost $4.62 billion on $431 million of revenue — a wider loss than a year ago, and roughly $97 billion of cumulative operating losses since 2019. Our valuation charges Meta harder than we charged Microsoft a day earlier, deliberately: Microsoft sells AI capacity to contracted customers, Meta's capacity serves Meta. We charge $45 billion a year of maintenance capex, expense the stock compensation, and still get about $80 billion of owner earnings — a probability-weighted fair value near $610 versus roughly $527 in the July 30 pre-market (the $585.61 pre-print close is the wrong basis; this reported after the close). At $527 the market is paying about 16.8x owner earnings, implying just ~6% annual growth for a decade from a company growing revenue 28%. Our call: BUY, 3 out of 5 — a modest 16% margin of safety, not a screaming one. Wall Street's aggregate target is $724.50 with a Buy consensus, but every single analyst who updated after this print cut their target — JPMorgan to $640, UBS $715, Goldman $725, TD Cowen $750, Citi $800, BofA $810, a median near $740. We agree on direction and differ hard on value: our $610 is below the lowest fresh target on the Street. Watch price per ad by region and 2027 capex guidance — those two lines decide everything. Not financial advice. THE CALL: BUY (3/5, WE VALUE IT BELOW EVERY FRESH STREET TARGET — AND IT IS STILL CHEAP) — base-case value ~$610.00 vs ~$527.34 today. What to watch: 2027 capital expenditure guidance coming in at or below 2026's level, free cash flow inflecting back above $10B a quarter, and the buyback restarting would each signal that the spending peak is behind Meta and would justify a higher fair value and an upgrade; the risks to respect are the mirror image — 2027 capex guided above roughly $170 billion, which means depreciation is chasing a target that keeps moving and never catches up, or US & Canada price per ad decelerating toward the Asia-Pacific pattern of flat pricing on more inventory, which would mean the AI spend is not actually buying pricing power at all Also on YouTube: @ChargedAlpha DISCLAIMER: For informational and educational purposes only. Not financial advice. Do your own research before any investment decision.