Mark Zuckerberg's net worth declined $18 billion in a single session after Meta Platforms posted second-quarter results showing free cash flow collapsed 91% year-over-year. The stock fell 10% as management outlined a $145 billion multi-year artificial intelligence infrastructure build with no clear timeline for positive unit economics. The move erased roughly $80 billion in enterprise value and marked the largest single-day wealth destruction for a tech founder since Elon Musk's $16 billion loss in November 2021.
Meta generated $1.8 billion in free cash flow during Q2, down from $20.3 billion in the prior-year period. Capital expenditures reached $14.2 billion for the quarter, a 112% increase year-over-year, driven entirely by data center construction and proprietary silicon development for large language model training. Management guided full-year capex to $37-40 billion, then disclosed a three-year envelope approaching $145 billion with no announced depreciation schedule or revenue attribution model. The disclosure arrived without the margin improvement narrative that typically accompanies infrastructure cycles.
The wealth erosion matters because Meta now operates in a regime where cash generation lags capital deployment by an order of magnitude, and the company has not articulated when—or whether—AI infrastructure becomes margin-accretive. Buyback capacity contracts when free cash flow approaches zero. Meta spent $14.5 billion on share repurchases in Q2, funded largely by existing cash reserves rather than operational cash flow. That arithmetic does not scale if capex remains elevated and revenue growth decelerates. Family offices holding Meta as a duration-adjusted growth position now face a binary outcome: either the AI spend produces incremental high-margin revenue within 18-24 months, or the stock re-rates downward as a capital-intensive utility with exposure to cyclical advertising demand.
The second-order effect is contagion risk across the Magnificent Seven complex. Alphabet, Amazon, and Microsoft have collectively guided to over $200 billion in combined AI-related capex through 2026, none with disclosed ROI frameworks. Meta's sell-off signals that public market investors will no longer underwrite open-ended infrastructure spend without clear monetization pathways or margin targets. If allocators demand proof of AI profitability from one hyperscaler, they will demand it from all hyperscalers. The repricing begins with the company that discloses the spending envelope without the revenue model.
Operators should monitor Meta's Q3 earnings call in late October for revised capex guidance or any early disclosure of AI-driven revenue contribution, particularly in WhatsApp Business API or Instagram shopping conversions. Watch for secondary sales by Zuckerberg under his pre-announced 10b5-1 plan, which could signal internal conviction around valuation. Allocators holding other hyperscalers with elevated AI capex should stress-test downside scenarios where free cash flow compression persists beyond 2025 without offsetting margin expansion.
The last time Meta's free cash flow fell below $2 billion quarterly was Q4 2018, during the post-Cambridge Analytica trust collapse. The stock required 14 months to recover its prior highs. This time, the contraction is self-inflicted and forward-committed.