Applied Materials, Lam, ASML on AI efficiency cycle

Applied Materials, Lam Research and ASML sit directly in that path. They are the toll roads on semiconductor capacity: every new wafer start, process node transition or yield improvement runs through them. Their businesses are tied to the capital intensity of chipmaking, and that is exactly why the current backdrop is powerful. AI remains a major driver of equipment spending, as both Applied Materials and Lam Research have said in recent filings, but the next round of demand is increasingly linked to customers trying to extract more output from existing footprints. That favors tools for deposition, etch, inspection, advanced packaging and process control — the parts of the chain that make each square meter of cleanroom more productive.

The price action says investors are already grappling with that transition. Lam Research has swung sharply, with the stock falling from above $430 in late June to about $277 on Oct. 19, while Applied Materials has slipped to around $423 from a June peak near $722. ASML has also pulled back from above $1,986 at the end of June to roughly $1,674. Technically, all three have weakened, with their shares trading below the 50-day moving average and momentum indicators softening. That does not change the structural story, but it does signal that the market is no longer willing to pay any price for AI exposure.

That is exactly where the asymmetric opportunity may lie. The market tends to think of semiconductor equipment as a single trade on capex growth. I think that is too simplistic. The better setup is to own the companies that benefit when customers cannot expand their footprint fast enough and instead must squeeze more chips, more yield and more automation out of every facility they already have. That is a much stickier, more durable revenue pool than pure greenfield construction.

Adalytica’s industrial production sentiment remains neutral, and PMI recession awareness is still elevated, which reinforces the same message: capital spending is not disappearing, but it is being rationed. In that environment, equipment vendors with deep technology moats, service revenue and exposure to process complexity should outperform lower-value hardware providers. The same logic extends beyond semis to factory automation names, industrial software and power infrastructure suppliers that help factories run denser, smarter and more efficiently.

For investors, the takeaway is simple: the AI factory cycle is not over, but it is becoming more selective. I believe the best way to play it now is not to chase every headline tied to chip demand, but to focus on the picks-and-shovels companies that enable production expansion without requiring a larger footprint. That is where the next multi-year return can be generated.