AI companies are forced to think about memory provisioning years in advance, not just when the next server arrives.
As models grow and workloads become more demanding, ensuring sufficient high-performance memory becomes an increasingly important operational and financial constraint. Bank of America calls this growing burden that of AI “memory tax”.
For Micron Technologies (MU), that means higher prices, longer customer commitments and a potentially more sustainable earnings cycle than investors have historically associated with memory chips.
In a note shared with me, Bank of America responded by sharply raising its guidance after another stunning profit, and the reason is far from simple improved results.
BofA says AI allows memory to capture more value, while Micron gains greater visibility through take-or-pay customer agreements.
Why Bank of America predicts a more sustainable Micron earnings cycle
What’s changed for Bank of America isn’t just that Micron has beaten expectations again, as it has for the last four consecutive quarters.
The bank now has better visibility into how long high prices and profitability can last, allowing it to increase its profit forecasts.
BofA pushed back its fiscal 2027 revenue estimate to $275.4 billion out of $230.3 billion and fiscal 2028 sales to $317 billion out of $244.1 billion.
More microns:
EPS estimates soared to $171.78 from $140.24 for 2027 and $197.90 from $145.45 for 2028. He also reiterated his Price target of $1,550while calling Micron one of its top picks in AI.
This belief comes from the fact that demand is becoming increasingly difficult to dismiss as a temporary spike in the memory cycle.
Micron now has 26 strategic customer agreements, up from 16 last quarter, covering approximately 35% of revenue through 2030. More than 75% of fiscal 2027 production is already committed under these agreements and other customer commitments.
CEO Sanjay Mehrotra went further, saying Micron had “no visibility” on when supply and demand would return to balance. He expects memory requirements in 2027 and 2028 to be stricter than in 2026, as larger AI models, longer context windows, and greater concurrency consume more memory.
Investors were initially hesitant after the earnings, with Micron falling about 4% before reversing to finish up about 3%.
This shows that even though Micron has killed it from an earnings outlook perspective, investors still need to decide how much of this AI-driven pricing power is truly structural rather than cyclical.
Gn bussni
