META Guides $145B Capex While Cutting 8,000 Jobs, Trading at Sub-1.0 PEG on 25% Implied Growth

Meta's May announcement of an 8,000-employee reduction (roughly 10% of workforce) combined with guidanceCompany-issued forecasts of future financial performance. for up to $145B in annual capital spending is not a contraction; it is a structural reorientation. The company is canceling 6,000 open roles and moving another 7,000 employees into AI-focused roles, a signal that headcount in legacy content moderation, ad operations, and non-core functions is being eliminated while AI infrastructure teams are expanding. This distinction matters because it reframes the narrative from 'austerity' to 'capital intensity pivot'.

Meta's capex guidanceCompany-issued forecasts of future financial performance. is now the highest of any mega-cap technology company on an absolute basis and tops even NVDA on a percentage-of-revenue basis. The spending is directed at training custom AI models (including large language models and multimodal systems that compete with OpenAI and Anthropic), building inference infrastructure to serve 3B+ users with AI-assisted features, and experimenting with non-public applications like neural interfaces and robotics that will define the next cycle. From a management perspective, this is a bet that proprietary AI will eventually drive more advertising upside than incremental content creators or engagement metrics.

Market reaction has been tepidly positive. Meta stock bounced to $614 (near ATH) after the announcement, and bear commentary on the job cuts has been muted relative to similar cuts at other tech firms. Analyst consensus seems to accept the reallocation as a sign of disciplined capital allocation rather than panic. The valuation implied by this capex trajectory is becoming cheap on legacy metrics: META trades at a PEG ratio below 0.9 (forward PE / growth rate) when normalized for the AI capex intensity, making it arguably the best-valued mega-cap.

The execution risk is significant. If Meta's custom AI models fail to deliver productivity gains or advertising upside within 18-24 months, the capex becomes a stranded asset and the workforce cuts become a permanent impairment. Competitors like Google have also bet heavily on AI infrastructure (through both Anthropic stakes and internal model development), so Meta's moatA sustainable competitive advantage that protects long-term returns on capital. is not assured. If OpenAI or Anthropic reach AGI-level performance first, or if their APIs become the default for enterprise AI applications, Meta's proprietary models could become secondhand assets with limited monetization paths. For now, the market is giving Meta credit for discipline and forward-thinking; if capex does not correlate with user engagement and ad ROI by late 2026, sentiment could reverse sharply.