AI Chip Spending Is Projected to Hit $1.6 Trillion by 2030. Here Are the Stocks Positioned to Capture Much of It.

McKinsey forecasts semiconductor spending could rise to ~$1.6T by 2030 from $775B in 2024 (13% CAGR), driven by AI demand for leading-edge chips and HBM. Nvidia is cited with ~85% annual revenue growth and 211% profit growth in Q1 FY2027 and a valuation positioned as “the cheapest since 2019” (P/E ~31 vs S&P ~33), while ASML’s EUV monopoly underpins growth (Q1 revenue +13%, net income +17%) despite a higher P/E (~58). SK Hynix, with ~56% HBM share and a Nvidia-focused HBM partnership, shows extreme growth (Q1 revenue +199%, net income +398%) amid supply constraints, though the article flags potential HBM boom-bust risk.

Analysis

This is really a bottleneck trade, not a generic AI demand trade. The cleanest second-order winner is ASML because the monetization is embedded in the installed base: even if wafer demand pauses, the service/upgrade stream is stickier than chip ASPs. TSM also benefits as the constrained manufacturing node, but the bigger point is that the semiconductor profit pool is tilting upstream toward scarce capacity and process tools rather than toward every chip name with AI exposure.

SKHY is the highest-beta expression of the supply squeeze, but it is also the most cyclical. HBM is still a scarcity market, so near-term pricing power should stay extreme for 2-4 quarters; after that, capacity adds from rivals can turn a shortage into a margin problem very quickly. That makes the stock attractive on dips, but only if investors keep sizing disciplined and treat it like a cycle, not a secular annuity.

Contrarian take: the market may be overpaying for linear extrapolation of AI capex while underestimating qualification lags and demand digestion. If hyperscaler budgets merely normalize, the first thing to break is not unit volume but multiple support for the highest-duration names. NVDA is still the quality anchor, but at this stage it is more vulnerable to an earnings or guidance reset than the supply-constrained names. The key falsifier is any evidence that order growth, lead times, or HBM pricing are rolling over in the next 1-2 quarters.

Key Decisions for Investors

• Prefer long ASML vs long NVDA on a 3-6 month view; ASML captures the bottleneck with more durable service economics and less competition risk. Add on any broad semi pullback; reduce if ASML bookings/gross margin soften.
• Initiate a modest long SKHY position only after post-listing liquidity stabilizes; treat it as a 2-4 quarter scarcity trade, not a long-term compounder. Exit/review if HBM lead times compress or pricing turns down.
• Pair trade: long TSM / short INTC over 6-12 months. TSM monetizes advanced-node scarcity; INTC faces the wrong mix of capex intensity and execution risk. Thesis breaks if INTC shows sustained foundry share gains or TSM capex discipline weakens.
• Do not chase MU here without confirmation that HBM share gains are translating into pricing power; if you want memory exposure, wait for next earnings to see whether the cycle is still in shortage mode.