From locking orders in HBM to endless demand for DRAM and NAND, Micron's (MU.US) "peak moment" is far from over.

The US storage chip giant Micron Technology (MU.US) will announce its performance for the second quarter of the 2026 fiscal year ending in February next week. Wall Street analysts generally believe that under the impetus of cloud computing companies and other technology giants locking in HBM capacity in advance on a large scale, as well as the incredibly strong demand for DRAM/NAND storage products in artificial intelligence data centers, the prices of storage products will continue to soar. The overall performance of this fiscal quarter and the future outlook of the management are expected to greatly exceed market expectations, and will become the core catalyst leading the storage sector and AI computing industry leaders towards a new round of strong growth trajectory. With the outbreak of full-scale war between the US/Israel and Iran spreading to multiple countries in the Middle East, igniting a new political superstorm of GEO Group Inc that is sweeping the global economy, investors' risk appetite has sharply cooled against the backdrop of skyrocketing oil and gas prices. Their serious concerns about the global economy, which is still in a fragile recovery process, are facing the severe threat of "stagflation" due to the out-of-control surge in energy prices, leading to recent severe losses in global stock and bond markets, as well as the cryptocurrency market. However, a recent research report released by the analyst team at Bank of America Corp on Wall Street stated that the latest supply chain surveys and storage industry tracking investigations show that the global storage industry centered around storage chips is still in a "super cycle". The impact of the Middle East's GEO Group Inc political conflict on the storage supply chain and the impact on fund managers' bullish outlook on the storage sector is almost negligible. Without a doubt, storage chips have become the most valuable asset in the AI era, and Micron is embracing the "AI super dividend". Whether it's the incredibly large TPU AI computing cluster led by Alphabet Inc. Class C, or the massive NVIDIA Corporation AI GPU computing cluster, both rely on HBM storage systems integrated with AI chips. In addition to HBM, current tech giants like Alphabet Inc. Class C and OpenAI are accelerating the construction or expansion of AI data centers and also need to purchase large quantities of server-level DDR5 storage and enterprise-level high-performance SSD/HDD storage solutions. Unlike Seagate and Western Digital Corporation, who focus on monopolizing near-line high-capacity HDDs, SanDisk focuses on high-performance eSSDs, and the three major storage chip manufacturers Samsung Electronics, SK Hynix, and Micron are currently leading in multiple core storage areas: HBM, server DRAM (including DDR5/LPDDR5X), and high-end data center enterprise-level SSDs (eSSD). They are the most direct beneficiaries in the "AI memory + storage stack" and can be said to collectively reap the "super dividend" of AI infrastructure. Micron is set to release its second-quarter earnings for the 2026 fiscal year after the US stock market closes on March 18th. The mainstream consensus on Wall Street for this earnings report is extremely high: revenue is expected to be around $19.15 billion, and non-GAAP adjusted EPS is expected to be in the range of $8.5 to $8.59. Based on the guidance provided by the company's management for the previous quarter (Q1), with revenue midpoint at $18.7 billion and adjusted EPS midpoint at $8.42, it means the market expects Micron to at least meet the upper end of the guidance, with some analysts even expecting slightly higher than the upper end outlook given by Micron's management team. Comparing to the same period last year, Micron's actual performance for the second quarter of the 2025 fiscal year was: revenue of $8.053 billion, non-GAAP adjusted EPS of $1.56, and GAAP EPS of $1.41. According to current Wall Street consensus forecasts, this means that Micron's revenue for the second quarter is expected to increase by approximately 137.8% year-on-year, and non-GAAP EPS is expected to increase by about 444.9% to 450%. In other words, the market is not expecting "modest growth", but an explosive growth in the financial report, and analysts generally believe that under the impetus of the near-endless storage demand driven by the AI data center construction frenzy, it is likely to greatly exceed the consensus expectations. Citigroup, a financial giant on Wall Street, recently maintained a "buy" rating on Micron stock and raised the target price for the stock within the next 12 months from $385 to $430 before the company announced its quarterly earnings. The core logic behind this move is that the prices of DRAM/NAND are still rising in the face of the extremely strong demand for artificial intelligence, and they believe this trend is not yet over. As of Monday's US stock market close, Micron's stock price rose by 5.14% to $389.32. The analyst team at Citigroup, led by Atif Malik, stated, "We have raised our earnings expectations for the February quarter and future quarters, above market expectations, mainly due to the strong memory price trend since the beginning of the year. Citigroup's global storage product analyst Peter Lee's forecast model shows that driven by the strong demand for data centers, the average selling price of DRAM in 2026 is expected to increase by 171% year-on-year; driven by the strong demand for eSSD, the average selling price of NAND flash memory is expected to increase by 127% year-on-year. In addition, media reports have indicated that Samsung will raise DRAM prices by 100% in the first quarter." BNP Paribas recently published a research report stating that the contract price of DRAM is expected to rise by 90% in the first quarter of 2026, with the long-term stable price curve of NAND expected to increase significantly by 55%, and prices are expected to continue rising in the second quarter since the latter half of 2025. The analyst team at BNP Paribas has even set a target price of up to $500 for Micron. BNP Paribas' assessment of the rising trend in storage prices is not an isolated view. TrendForce recently raised its expectations for the regular contract price of DRAM in the first quarter of 2026 from the previous estimate of a quarterly increase of 55%-60% to +90%-95% QoQ (quarter-on-quarter), and significantly raised the contract prices for NAND Flash by +55%-60% QoQ. The research also pointed out that North American cloud computing companies have seen a sharp increase in demand for enterprise SSDs, driving prices to rise by 53%-58% QoQ in the first calendar quarter. All of these factors indicate a critical point: storage chips have become a key bottleneck in the supply chain, realizing pricing power in the AI super wave that is not inferior to NVIDIA Corporation's AI chips. Multiple overseas media reports suggest that Samsung, the largest player in the storage industry, has raised the prices of dynamic random access memory (DRAM series storage products) by over 100%. According to Korean electronics news, Samsung has completed final negotiations on DRAM supply prices for the first quarter with major customers such as Apple Inc. The average price of server, PC, and mobile-grade DRAM has increased by about 100% from the previous quarter and doubled compared to the fourth quarter of last year, with some customers and products seeing price hikes of over 100%. The report cited insider sources revealing that the negotiations have been fully concluded, and some overseas customers have completed payments. This increase is higher than the 70% level negotiated in January, expanding by approximately 30 percentage points in just one month. The super cycle in storage is ongoing - this is not just an ordinary financial report, but a verification of a "price surge". The rapid escalation of DRAM/NAND series storage product prices is reshaping the long-standing contract conventions in the global storage industry, especially with the increasing reliance of GPU/TPU systems on HBM, DRAM, and enterprise-level SSDs causing a long-term imbalance in supply and demand. The negotiation cycle has compressed from traditional annual contracts to quarterly contracts, and now even needs to be adjusted monthly, reflecting the severity of the supply-demand imbalance in the storage chip market. From a fundamental hardware perspective, AI computing is not only limited by computational power but also by "data transportation capability". The efficiency of large model training and inference is determined not only by the quantity of Tensor Cores/matrix units, but by how quickly weights, KV caches, activation values, and intermediate tensors can be fed into the computation cores per second. From the perspective of the intersection of semiconductor and AI data center infrastructure, storage chips perfect their position in the AI wave not only because they benefit from the expansion of training and inference but also because they are a universal toll booth across platforms, architectures, and ecosystems. As the AI era transitions from training dominance to inference, agents, long contexts, enhanced retrieval, the demand for capacity, bandwidth, power efficiency, and persistent data layer only grows stronger. The storage systems that AI data centers heavily rely on are not just HBM. The complete AI storage hierarchy includes: HBM responsible for high-speed supply to the closest accelerators, DDR5/RDIMM/LPDRAM responsible for memory expansion and data preprocessing in the host, and enterprise-level SSDs responsible for training datasets, checkpoints, vector libraries, RAG retrieval, and inference caching for persistent data pathways. For example, Micron defines its AI data center storage solutions as a complete combination of storage devices covering training and inference; it also explicitly states that its eSSD product line is used to maintain efficient data supply throughout the AI pipeline for both training and inference. TrendForce also points out that with the advent of the AI inference era, North American cloud computing giants are rapidly increasing their purchases of high-performance storage, with demand for eSSDs exceeding expectations. This means that AI GPU clusters cannot operate without storage, and Google's TPU clusters are similarly dependent on storage the only difference is the accelerator brand, but the underlying storage layer must be built on the complete pyramid of HBM + server DRAM + NAND/SSD. What's more, the price hikes are not over yet. Industry chain sources from DigiTimes indicate that DRAM prices in the second quarter of 2026 may increase by about 70%; at the same time, Phison (a subsidiary of United Microelectronics Corp.) has begun discussing prepayment arrangements with customers as NAND prices continue to soar and supply remains tight. This type of transaction structure of "prepayment first, then lock in goods" fundamentally illustrates that the market has shifted from a regular purchasing mode to a seller's market with a "grab resources first" approach. Tom's Hardware's summary of this round of the super cycle in storage is more straightforward: DRAM prices are even beginning to show characteristics of "changing by the hour". The spot market for DRAM can further illustrate that the super cycle in storage is far from over. TrendForce data from early March shows that the average price of DDR4 1Gx8 3200 in the spot market rose from $32.34 to $33.02 in just one week, while the price of 512Gb TLC wafers increased by 14.7% in a single week, breaking through $20.586. This indicates that not only contract prices are high, but spot prices are still on the rise; and when spot prices are higher than contract prices, it usually means there's pressure for future contract renegotiation to increase. For a company like Micron that is simultaneously exposed in HBM, DRAM, NAND, and enterprise-level SSD product lines, this is the classic sign of favorable average selling prices. Locking in bulk orders of HBM in advance and soaring prices of DRAM/NAND Micron's management had clearly stated at the end of last year that the supply of HBM for the 2026 calendar year had already been secured with agreements on price and quantity, covering products including HBM4; additionally, the company stated that the tight balance formed by strong demand and supply constraints is expected to continue at least until after 2026. This means that the most profitable and scarce part of Micron's business, the revenue and profit base, is already locked in. This is why the market is not just looking at whether there will be an outperformance, but how much it will exceed and if there will be an upward adjustment in the annual performance outlook. Therefore, the Wall Street analysts focusing on the bullish prospects for Micron are not just waiting for a strong financial report, but for the management to affirm the super cycle in storage ignited by AI is far from over. Micron is not now a storage chip company with strong performance, but the entire storage chip industry is lifting it up. Micron's strong HBM revenue for the 2026 calendar year has been locked in advance, and is actively securing large orders from cloud computing giants for 2027, while the non-HBM side is enjoying the overshooting benefits of price hikes. The company's performance structure is shifting from "betting on supply-demand cycles" to "securing advanced high-end HBM orders + benefiting from comprehensive price hikes". Therefore, in the view of some analysts, the current market may be underestimating the impact of explosive growth in average selling prices on Micron's performance outlook for the second half of 2026. If we shift the perspective from just a single quarter's financial report to a yearly view, the core logic driving the Micron stock price bull market is actually the "revaluation of the classic storage cycle by AI" this is concentrated in the 240% increase in the Micron stock price for the full year of 2025, with the increase already reaching 40% so far in 2026. There's currently no sign of a downturn in the "storage super cycle", but rather the AI data center construction frenzy continues to push the demand for HBM, enterprise-level SSDs, and server-level high-performance DRAM to levels far exceeding the industry's annual maximum capacity. This represents a slightly conservative forecast for performance upgrades, but for investors looking at Micron's future stock price trajectory, it's quite exciting: if Micron's EPS for the 2027 fiscal year continues to increase by 10%-15% within 2026 (the average Wall Street estimate), and given a median P/E ratio of 11.5, then the target price could reach $578.6 per share, representing over 40% upside from the current strong stock price. However, whether this figure will ultimately be achieved depends on two things: firstly, whether the rising price trend and supply-demand imbalance will continue to transmit well into the first half of 2027 or even the full year; secondly, whether the management will clearly communicate to the market that this scarcity will not rapidly reverse in the coming quarters. Based on current signals from the storage industry chain, the long side seems to have a strong foundation to support it.