Bloomberg Surveillance · Oct 22, 2026
With Cameron Dawson, CIO of NewEdge Wealth
Semiconductor earnings have grown more than 100% so far this year — and Wall Street already expects that pace to fall by half in 2027. Most strategists would call a deceleration to 50% growth a strong year by any normal standard. Cameron Dawson, chief investment officer at NewEdge Wealth, told Bloomberg Surveillance that the market is instead treating it as proof that 2027 is peak earnings for chip stocks, without AI spending even slowing down.
"We don't have to stop the spending spree to see AI infrastructure earnings stumble." — Cameron Dawson
Cameron's own note on the semiconductor trade was read on air by host Jonathan Ferro before she even started answering questions, so the whole segment ran as a defense of that note rather than a fresh set of questions.
The full segment is covered here so you can skip it.
Here are the 4 takeaways that matter.
Key Takeaways
Chip-earnings growth is set to roughly halve in 2027 — from over 100% this year to about 50% — even with nobody calling for AI spending to stop
Semiconductor stocks trade on counter-cyclical multiples, so the cheapest valuation lands on the highest earnings right as the market prices in the peak
This year's chip-earnings beat came from margin expansion, not just more spending — the SOX index's operating margin jumped from 16% to 42%
In every past technology cycle, Cameron said, the stocks peaked before adoption did — and she expects 2027 to repeat that pattern
Recent improvement in market breadth is misleading, since it's being driven by a rotation into defensives, not cyclicals
1. The Spending Won't Stop
Cameron's note set the tone for the whole segment: a stumble in AI infrastructure earnings doesn't require AI spending to actually slow down.
AI infrastructure earnings can wobble even while spending keeps rising
"We don't have to stop the spending spree to see AI infrastructure earnings stumble." — Cameron Dawson
The distinction matters for anyone waiting for slower spending as the signal to watch. Cameron's argument is that stock prices, not spending plans, are what is already pricing in a peak — and that is an earlier, separate signal.
