BofA Strategist: AI Capital Returns in Question, Defensive Sectors Show Rotation Potential
N.R. Finch
BofA European equity strategist Sebastian Raedler warns that AI's pricing power is eroding — rising prices push clients to cut usage, open-source models undercut margins — and if a capex inflection hits, defensive sectors become the next pricing anchor.
AI capex is pouring out — who actually pays the bill?
Raedler's core question: the massive AI capex driving demand for Micron and other chipmakers must ultimately be recouped from end users paying for AI services.
This means → the profit foundation of the entire chain sits not with suppliers (chips, cloud) but with how much end users are willing to pay.
He calls current earnings forecasts "very optimistic" — the market prices in sustained high-price adoption, an assumption reality is already chipping away at.
Why is AI's pricing power slipping?
Raedler identifies two forces: price hikes are pushing clients away — as AI services raise prices, some users simply cut usage.
At the same time, low-cost open-source models are eroding the scarcity that closed-source providers need to maintain fat margins.
In plain terms = customers are balking at the price on one side; cheap alternatives are flooding in on the other — both squeeze the margin moat narrower.
What signal would crack the AI trade?
Raedler draws three trigger lines: ① clients explicitly refuse to pay up; ② traffic shifts en masse to cheaper models; ③ the first hyperscaler admits "I'm not sure I'm creating value" and cuts capex.
This means → the three need not fire together — any single one shakes the market's faith in AI capital returns.
This reflects how fragile the current AI trade is: it rests on the expectation that spending will convert to profit, not on verified cash flows.
Why can defensive sectors absorb the rotation?
Consumer staples and pharma are what Raedler calls the "forgotten sectors" — capital chased AI for years, compressing their valuations, yet their fundamental logic remains intact.
In plain terms = regardless of where AI goes, people still buy medicine and daily essentials — these sectors' revenue does not depend on the AI narrative.
He adds that if investors judge AI capex unsustainable, German bunds and various pair trades should also outperform ahead of the inflection point.
How should we read the market's mood right now?
The Nasdaq 100 fell 3.3% on Tuesday; upbeat guidance from Micron and Qualcomm then lifted semiconductor sentiment.
This means → near-term data still props up the AI-demand narrative, but Raedler's logic points further out: if this recovery proves unsustainable, defensive sectors' relative value becomes the next pricing anchor.
This reflects a market sitting at the "faith vs. verification" tipping point — each recovery demands stronger evidence to hold.
Content is for reference only, not financial advice.