Micron Drops 7% Premarket as Analyst Warns Memory ASPs Could Peak by Mid-2026
Miles Bennett
Micron Technology fell 7% pre-market to $1,005, after Raymond James analyst Karl Ackerman forecast that DRAM and NAND average selling prices will peak by mid-2026 — roughly a year earlier than Wall Street consensus, raising the question of whether memory's AI rally is shorter than the market has priced in.
What is the market actually afraid of?
The surface trigger was broad semiconductor weakness, but the deeper pressure is a shifting view on when the memory price cycle tops out.
Raymond James analyst Karl Ackerman stated in a Wednesday report: DRAM and NAND ASPs will peak by mid-2026.
This means → the consensus thesis — "AI demand supports pricing through mid-2027" — just lost a full year.
In plain terms = the window of rising memory prices may close sooner than most investors expected.
Why does this call diverge from Wall Street consensus?
Most analysts had placed the point at which supply growth materially pressures demand at mid-2027 or later.
Their core logic: AI infrastructure buildout will sustain structural demand for products like HBM — high-bandwidth memory designed for AI accelerators.
Ackerman's divergence: he sees supply catching up faster, pulling the impact window forward by roughly a year.
This reflects a market narrative shifting from uniform optimism on AI demand duration toward genuine disagreement.
Can long-term contracts cushion the blow?
Ackerman did not downgrade Micron — he maintains an "outperform" rating.
His reasoning: long-term pricing agreements with Google, Meta, and other hyperscalers will partially buffer the revenue hit from falling spot prices.
In plain terms = even if market prices drop, contract prices hold for a while, so this downturn should be milder than past memory cycles.
The key question remains: can contracts lock in the price floor, or merely delay the decline — those are two very different outcomes.
How much is already priced in?
Micron's forward P/E has expanded from 4.4x in April to 11.7x (per FactSet).
This means → the stock has already front-loaded a significant portion of the "continued high growth" expectation.
Ackerman argues this multiple embeds three pressures: slowing contract ASP growth + margin compression + gradual oversupply forming over the next one to two years.
Put simply = the stock is paying for the most optimistic script — if reality falls even slightly short, the correction room is material.
Korean rivals are falling too — what does that signal?
SK Hynix and Samsung Electronics fell 2.63% and 2.50% respectively on the same day.
This reflects a sentiment shift targeting not just Micron but the entire memory sector.
Over the past year, Micron's stock rose roughly 916%, with its market cap briefly crossing $1 trillion — all driven by the AI demand thesis.
The real unresolved question: whether 11.7x forward P/E can be validated by earnings over the next two years — if ASPs peak earlier than expected, the validation window becomes very tight.
Content is for reference only, not financial advice.