The AI semiconductor trade remains "remarkably resilient" despite rising interest rates, according to the head of semiconductor research interviewed by Bloomberg — a striking verdict in a year when higher borrowing costs have battered nearly every other corner of the tech sector.

That phrase, "remarkably resilient," is doing a lot of work. It comes from a research head whose job is to track order books, capacity plans and pricing across the chip supply chain, not to talk up the market. And it lands at an awkward moment for macro-driven investors.

Why does this matter now?

Central banks have spent the past two years pushing rates upward to tame inflation. Higher rates normally squeeze the long-duration growth trades first: software, unprofitable startups, and semiconductors priced on earnings years away. Chip stocks have lived through that squeeze more than once.

Yet the AI segment of the trade has refused to break. The demand signal from datacenter buildouts has proven strong enough to override the financing headwind — at least so far, in the assessment Bloomberg's interview subject shared.

That resilience has a concrete commercial meaning for the supply chain. When chip demand holds firm through a rate cycle, fabs keep ordering tools, foundries keep pricing tight on advanced nodes, and suppliers of AI accelerators and high-bandwidth memory keep their leverage in negotiations. The usual cyclical relief buyers hope for — pricing concessions as demand cools — hasn't materialized in the AI-exposed part of the market.

What does "resilient" actually mean here?

It's worth separating what the research head confirmed from what remains a view on the road ahead. The confirmed part: AI-related semiconductor demand has held up under monetary tightening that would, in past cycles, have dented it badly.

The open question: how long that lasts. Resilience is not immunity. If rates stay high for longer, capital budgets eventually get repriced, and even AI infrastructure projects — with their multi-year payback profiles — sit exposed to the cost of money. The research head's framing, "remarkably resilient," carries an implicit surprise: the observer expected more damage than has shown up.

The context that changes the picture

Geopolitics and supply concentration matter here because they reinforce the demand story rather than competing with it. Governments in the US, Europe and Asia have committed public money to domestic chip capacity, and those programs add a floor under equipment and construction spending that doesn't move with the rate cycle. AI demand and industrial policy are pushing in the same direction.

For chipmakers, that combination has kept order visibility long. For buyers, it means the bargaining power imbalance of the past two years — sellers dictating lead times and allocations on AI-relevant silicon — persists into a rate environment that history says should have corrected it.

The Bloomberg interview is, ultimately, a snapshot of a market defying its own macro script. Whether "remarkably resilient" becomes "remarkably durable" depends on whether AI demand keeps justifying its capital intensity once the full cost of money is finally felt across the industry's ledgers.