Meta and Amazon are leading a trillion-dollar Big Tech spending spree

Morgan Stanley projects Big Tech capex will hit new record highs as AI infrastructure costs keep rising, with spending led by Meta and Amazon and also involving Alphabet, Microsoft, and SpaceX. The AI race is expected to intensify as supply-chain bottlenecks force companies to add more capacity. Investors can take comfort that pullbacks in AI spending appear unlikely, supporting a constructive near-term view of mega-cap AI capex momentum.

Analysis

This is less a growth story than a capital-intensity re-rating. The incremental dollar of AI spend is increasingly flowing to the plumbing of compute rather than to software-like margin expansion, which should keep pricing power strong for networking, power, cooling and semiconductor capacity. The hyperscalers still win strategically, but their equity holders may not: higher depreciation and utility expense can cap free-cash-flow yields even when revenue is healthy, so the market is likely to reward the enabling supply chain more than the balance sheets funding it.

The near-term risk is a digestion phase over the next 1-3 earnings cycles. If capex growth continues to outrun revenue growth, investors will start demanding proof that AI dollars are converting into incremental monetization rather than just preserving share. That would pressure the highest-spend names first, especially where the payback is least visible, and could trigger multiple compression even without any fundamental slowdown. The reversal signal would be any guidance shift toward flattening capex, shorter lead times, or improving return-on-invested-capital metrics before usage growth is durable.

The consensus is probably underestimating how much scarcity itself helps incumbents. Bottlenecks make it harder for smaller cloud and AI entrants to catch up, so aggressive spending can actually widen the moat of the largest platforms over 6-18 months. The contrarian trade is to separate the winners in the supply chain from the funders of the race: the former should see order-book leverage first, while the latter may need a monetization surprise to justify continued spending at current multiples.

Key Decisions for Investors

• Long SMH / short XLK for 1-3 months: prefer the infrastructure beneficiaries over the hyperscaler-heavy index; thesis breaks if AI capex guidance decelerates faster than expected or if software monetization reaccelerates.
• Build a long VRT and ANET basket on pullbacks for 6-12 months: both are direct levered plays on data-center buildout and should compound if bottlenecks keep pricing firm; watch for order backlog inflection as the key catalyst.
• Pair trade: long MSFT / short META into the next earnings cycle. MSFT has the cleaner enterprise monetization path, while META carries more capex intensity relative to near-term proof of payback; exit if META guides FCF materially better than expected.
• Use a META put spread 1-2 earnings out only if capex guidance is revised up again without a matching revenue lift; this is a tactical hedge against multiple compression, not a structural short.