Micron's AI Memory Momentum Forecast to Endure for Years

Instructions

Leading financial analysts are increasingly optimistic about Micron Technology's prospects, driven by the escalating demand for artificial intelligence (AI) memory solutions and sustained strong pricing in the memory market. This positive sentiment suggests a prolonged period of growth for the company, with projections indicating substantial earnings expansions fueled by innovations like high-bandwidth memory (HBM) and robust data center requirements.

Krish Sankar, an analyst at TD Cowen, has reiterated a 'Buy' rating for Micron, elevating his price target from $450 to $500. His revised estimates are based on the expectation of Micron surpassing earnings forecasts, eyeing a valuation multiple of 5-6 times the projected 2027 earnings. Sankar acknowledges the competitive landscape, with Micron already a popular long-term investment, yet he foresees further upside potential. However, he emphasizes that long-term stock appreciation will hinge more on a re-evaluation of the company's valuation rather than short-term earnings surprises. He forecasts impressive earnings per share (EPS) figures: $10.40 for the February quarter (exceeding the Street's $8.82 estimate) and approximately $13.50 for the May quarter (compared to the Street's $10.92 estimate). For 2026 and 2027, Sankar models EPS of $65 and $90, respectively, with book value per share potentially reaching $215 by the end of 2027. He also notes that strategic long-term agreements (LTAs) could play a crucial role in establishing a floor for gross margins beyond 2027, drawing parallels to the valuation multiples seen in the hard disk drive industry.

Similarly, Srini Pajjuri of RBC Capital has maintained an 'Outperform' rating for Micron, increasing his price target from $425 to $525. Pajjuri's upgraded outlook is anchored in the continued strength of memory pricing and the insatiable demand emanating from the AI and data center sectors. He projects a sharp increase in DRAM contract pricing, with TrendForce anticipating an 80-85% surge in the first calendar quarter. His models predict over 70% blended DRAM pricing growth and approximately 80% for NAND in the February 2026 quarter, leading to a revised revenue forecast of $25.5 billion and EPS of $13.52 for the quarter. For the May 2026 quarter, Pajjuri expects a 20% blended increase in DRAM pricing and a 30% sequential rise in NAND pricing, translating to an estimated revenue of $31.3 billion and EPS of $17.19. He has also raised his full-year EPS estimates to $54.30 for 2026 and $75.44 for 2027.

Both analysts concur that the burgeoning demand from AI and data centers will extend the current memory upcycle well into 2027. Data centers now account for more than half of the industry's DRAM revenue, and the robust demand for high-bandwidth memory (HBM) and DDR memory in AI systems is expected to persist. Micron's management has confirmed that HBM supply for 2026 is already fully committed, with HBM4 projected to commence volume shipments in the first calendar quarter. While HBM3E pricing might see a modest decline in 2026, HBM4 pricing could be 30-50% higher. The increasing memory capacity required by next-generation AI hardware, such as Rubin Ultra GPUs demanding significantly more HBM, further underscores this trend. This AI-driven demand is expected to more than compensate for any potential softening in the PC and smartphone markets, thereby sustaining strong growth across both DRAM and NAND throughout the upcoming market cycle.

In conclusion, the semiconductor industry, particularly in the memory sector, is poised for sustained expansion. The analytical community anticipates that Micron Technology will capitalize on the robust demand for AI-driven memory solutions, leading to impressive financial performance. While the market may eventually demand a re-rating of Micron's valuation, the underlying strength in pricing, coupled with innovative memory technologies like HBM, presents a compelling growth narrative for the foreseeable future.

READ MORE

Recommend

All