Crushing Wall Street’s Biggest Doubt: Micron’s Record 84.9% Gross Margin Sets Tone: AI Memory Shortage to 2027

TradingKey - After the US market close on June 24, memory chip giant Micron Technology ( MU) reported its fiscal third-quarter 2026 financial results, completely reversing market pessimism regarding AI memory demand with explosive core metrics.

This historic earnings report not only dispelled concerns over an "AI demand slowdown" and "earnings peaking," but also proved through actual performance that the AI infrastructure investment boom continues to propel the memory industry into an unprecedented upcycle.

Micron's fiscal third-quarter revenue surged approximately 346% year-over-year to $41.46 billion, beating analyst estimates by about 16%. Non-GAAP adjusted earnings per share rose more than 12-fold year-over-year to $25.11, exceeding expectations by over 20%. Net income reached $28.2 billion, nearly a 14-fold increase from a year ago and 16.5% higher than expected. The adjusted gross margin for the fiscal third quarter reached 84.9%, more than double that of a year ago and also beating expectations, indicating that memory chip pricing power remains extremely strong.

Following the earnings release, Micron's shares, which closed down nearly 0.4% during the day, surged into positive territory, with after-hours gains exceeding 16% at one point.

Micron management also noted that driven by the ongoing explosion in AI training and inference demand, tight HBM supply is expected to persist beyond 2027, and the company is further expanding its advanced memory and packaging capacity to meet this demand.

Record 84.9% gross margin locks in long-term dividends

In Micron Technology's fiscal third quarter 2026 earnings report, an adjusted gross margin of 84.9% emerged as the most eye-catching core metric. This figure not only set a record high for the company but also surpassed tech giants such as Nvidia ( NVDA) and Meta ( META ), ranking at the top among major U.S. technology companies.

"Our fiscal third-quarter gross margin more than doubled from the same period last year, an achievement sufficient to define a new era for the memory industry," Mark Murphy, Micron's Chief Financial Officer, said during the earnings call.

Looking at the data, the surge in Micron's gross margin was no accident.

In the fiscal third quarter, DRAM revenue increased by 67% quarter-on-quarter, but shipment bit growth was only in the single digits, with prices rising about 60% quarter-on-quarter; NAND flash prices rose by approximately 85% quarter-on-quarter. Behind this simultaneous growth in volume and price is the sharp contradiction between AI-driven memory demand and supply bottlenecks.

Tech giants such as Nvidia, AMD ( AMD) and Google ( GOOGL) require Micron's High Bandwidth Memory (HBM) to support their AI processors. Data center operators are snapping up memory regardless of cost, and even consumer electronics manufacturers like Apple are facing pressure from rising memory costs. Apple CEO Tim Cook noted last week that the iPhone would have to raise prices to cope with "unsustainable" memory pricing.

Micron's gross margin level has far exceeded conventional industry perceptions. In a traditional memory cycle, such a high gross margin typically indicates that the industry is in a phase of severe undersupply, but Micron's management has hinted that this trend may last longer.

The company expects its fiscal fourth-quarter gross margin to rise further to approximately 86%, and revealed that it has signed "Strategic Customer Agreements" (SCAs) with core customers to lock in future long-term supply relationships and pricing levels. This shift in business model signals that the memory industry is transitioning from a traditional cyclical sector to one of long-term stable growth.

Micron CEO Sanjay Mehrotra stated that due to AI-driven demand across various market segments and structural supply constraints, the supply tightness will persist past 2027, with no clear timeline yet in sight for supply to catch up with demand.

Data center revenue beat expectations by 69%

The explosive growth in the data center business directly addresses market doubts regarding the sustainability of AI demand.

Revenue from the core data center business reached $11.52 billion, beating analyst estimates of $6.8 billion by approximately 69%—a truly stunning margin of outperformance.

More critically, this data thoroughly dispels market concerns over cooling demand for AI server memory. The robust performance of the data center segment demonstrates that demand driven by AI training, inference, server upgrades, and high-performance storage is not slowing down, but is instead accelerating.

Meanwhile, the cloud storage business also delivered stellar results, posting revenue of $13.77 billion, which beat expectations of $10.69 billion by about 29%. This indicates that AI-driven storage demand is spreading beyond the isolated AI accelerator chain into broader data center storage and cloud infrastructure investments.

Micron Signs 16 Strategic Long-Term Agreements, Locking In $100 Billion in Revenue

Micron disclosed in its financial report that it has signed 16 "Strategic Customer Agreements" (SCAs) with key customers, including data center operators and automakers, covering approximately 20% of its DRAM shipments and about one-third of its NAND shipments, with agreement terms generally spanning from 2026 to 2030.

Most of these agreements adopt a "take-or-pay" model, meaning customers must pay for the contracted volumes regardless of whether they take delivery, providing Micron with unprecedented revenue certainty.

Under accounting standards, Micron's first-ever disclosure of its remaining performance obligations (RPO) shows that the signed agreements represent approximately $100 billion in guaranteed revenue. Meanwhile, the company is also set to receive about $22 billion in cash deposits and related financial commitments, with approximately $18 billion as cash deposits.

The core value of this strategy lies in locking in more than half of the company's revenue along with a significant portion of prices and volumes in advance, overlaying a stable layer of long-term contract revenue on top of the traditional spot business.

Micron CEO Sanjay Mehrotra emphasized that the price floors established in the agreements enable the company to achieve strong gross margins that far exceed the quarterly peaks of any previous cycle. Management specifically pointed out that even at the minimum price levels stipulated in the agreements, Micron's gross margin remains higher than the peak of any past cycle.

Micron's Q4 Guidance Far Exceeds Expectations: AI Chip Supercycle Far From Over

While current-quarter results only validate past growth, next-quarter guidance is the key to determining market valuation. Micron's latest fourth-quarter outlook not only far exceeded market consensus but also shattered external concerns that its earnings had already peaked.

The company expects adjusted revenue for the fourth fiscal quarter to reach $49 billion to $51 billion, with a midpoint of approximately $50 billion, which is about 15.6% higher than the market consensus of $43.24 billion.

Even more notably, this guidance represents a quarter-on-quarter growth of approximately 20.6% compared to the actual revenue of $41.46 billion in the third fiscal quarter, reflecting sustained strength on the demand side.

On the earnings front, Micron's adjusted EPS guidance was equally impressive, expected to reach $30 to $32 with a midpoint of $31, which is about 22.5% higher than the market expectation of $25.31, and still offers further upside compared to the third fiscal quarter's $25.11.

During the earnings call, Mehrotra emphasized that AI has become one of the most significant growth drivers for the memory industry in decades, and this trend is still in its very early stages.

He pointed out that, driven by AI-fueled demand across various market segments and structural supply constraints, the tight supply-demand dynamics for DRAM and NAND are expected to persist beyond 2027, with supply only gradually improving in 2028, though "there is currently no telling when supply will catch up with demand." This assessment is more optimistic than previous market expectations, suggesting that this round of the AI infrastructure buildout cycle could last much longer than investors had originally anticipated.

To meet the growing demand for AI memory, Micron is accelerating its capacity expansion. The company has raised its capital expenditures for fiscal year 2026 to $27 billion, with a primary focus on advanced-node DRAM and HBM capacity.