Tech Stocks Can't Be Stopped. Has the Iran Decoupling Begun?

The market is treating AI as a secular earnings compounder again, but the bigger signal is dispersion: mega-cap tech is reasserting leadership while the rest of the market is still digesting macro uncertainty. That usually happens when investors conclude that incremental capex, model training, and inference demand will disproportionately accrue to the pick-and-shovel layer rather than the application layer. The cleanest beneficiaries are still the infrastructure names with pricing power and strategic bottlenecks, especially where capacity is constrained enough to sustain margins even if end-demand later normalizes.

Second-order, this rally is more about supply allocation than narrative. If hyperscalers and AI labs keep chasing compute at the current pace, the risk is not just higher revenue for NVDA and AVGO, but a gradual re-rating of the entire semiconductor stack as lead times tighten, utilization rises, and customers lock in longer-term supply commitments. INTC and AMD benefit from the read-through only if they can convert the enthusiasm into share gains or price discipline; otherwise, higher industry pricing can actually widen the gap between the perceived AI winners and the rest of the chip group.

The main contrarian issue is valuation endurance, not valuation level. A multiple of 37-47x is defensible only if earnings revisions keep moving up over the next 1-2 quarters; if the upcoming large-cap reports show capex discipline, slower cloud growth, or weaker margin guidance, this move can unwind quickly because positioning is likely crowded after a 15-session run. In that sense, the near-term catalyst is binary: strong guidance extends the trade, but even a modest miss would likely compress the most extended names first, with semis usually moving more violently than software on disappointment.