"Controversy over the arrival of the storage super cycle" Morgan Stanley removes SanDisk (SNDK.US) "preferred" investment label.

More and more people are beginning to accept the concept of the semiconductor storage super cycle, but controversy still exists. Recently, Morgan Stanley released a detailed research report on SanDisk, stating that SanDisk is a controversial stock, and analyzing the controversy from the perspectives of demand, supply, and valuation. The conclusion is that even after a significant increase in stock price, SanDisk is still bullish, but profit growth needs time to match the current stock price, so they are removing its "top pick" designation. Part.01 Controversy One: Next year's demand growth and its impact on shipments Morgan Stanley's view: Industry bit growth is expected to reach 20-25% next year, the highest level in recent years. In the baseline scenario, SanDisk will grow in sync, and in the optimistic scenario, with new QLC products and more flexible supply growth, SanDisk is expected to outperform the industry. In 2026, the contribution of AI and data centers to NAND industry growth will reach a turning point. It is certain that enterprise solid-state drive will grow by 40-50% year-on-year in 2026. Different institutions have differing forecasts for the size of the enterprise solid-state drive market, but combining the procurement volume of large-scale manufacturers (each major buyer is about 40EB), a bit growth increment of over 120EB in 2026 for enterprise solid-state drives is fully achievable. Part.02 Controversy Two: What is the outlook for NAND supply growth? Morgan Stanley's view: NAND supply growth will be limited in 2026, and the joint venture between SanDisk and Kioxia is one of the few potential new sources of wafer capacity, which will support a positive assessment of pricing for 2026. Various variables on the supply side are being monitored: Positive factors: Equipment manufacturers confirmed that investment in the first half of 2026 will be limited. Bit output growth requires additional spending, but semiconductor equipment supply chain companies have not seen a near-term turning point in the NAND field, with spending expected to recover concentrated in the second half of 2026 (corresponding to supply growth closer to 2027). Core NAND equipment supplier Lam Research's third-quarter NAND system revenue is down by about 30% quarter-on-quarter, and it is expected to gradually recover, but significant acceleration will have to wait until the second half of next year, in line with statements from Richardson Electronics, VAT, and large OEM manufacturers. DRAM investment remains a priority for Micron, SK Hynix, and Samsung, driven by HBM demand and NAND bit growth lower than pre-2022 levels. Despite improvements in NAND pricing, DRAM profitability remains significantly higher, and the additional demand for wafers from HBM will continue to drive investment towards the higher-return DRAM sector. Part.03 Controversy Three: Reasonable profit outlook and trading range for SanDisk Morgan Stanley's view: Profitability is expected to increase significantly, with a forecasted EPS of $16.35 in 2026, and each cycle peak EPS is expected to reach $30. In the baseline scenario, SanDisk is expected to grow by 15% in 2026, benefiting from the ramp-up of BiCS8 capacity, a 12% year-on-year decrease in unit bit costs (excluding costs for underutilized capacity and start-up costs, down by 8% year-on-year), a 14.4% year-on-year increase in pricing (no increase in the second half of 2026), supporting a gross margin of 45.7% and an EPS of $16.35. In the optimistic scenario, revenue growth is higher, with bit growth close to 25%, a year-on-year decrease in unit bit costs of 8% (driven by higher-cost enterprise solid-state drives boosting revenue growth), a 31% year-on-year increase in pricing, and a quarter-on-quarter increase in pricing in 2026 (12% in the first quarter, followed by 10%, 3%, 3% subsequently, and a 4% increase in the fourth quarter). Interest expenses are halved from the current estimate of $47 million per quarter, with no change in interest income. Ultimately, revenue is expected to reach $13.1 billion, with a gross margin of 50.3% and an EPS of $26.26 (non-GAAP). If the gross margin can reach the mid-50% range, EPS is expected to exceed $30. The product portfolio remains SanDisk's biggest concern in the short term, with high enthusiasm in the enterprise solid-state drive market but low exposure for SanDisk (only 12% bit share in the second quarter). SanDisk has lagged behind in the enterprise solid-state drive market for years, but excluding Yangtze Memory, its bit share has remained stable. However, pricing for enterprise solid-state drives is higher, so in years where growth is being driven by enterprise solid-state drives, SanDisk's average price/income growth may lag behind its peers. In the optimistic scenario, shortages in hard drives and rising demand for enterprise solid-state drives will accelerate the certification cycle, with SanDisk leading the way. With the ramp-up of BiCS8 node capacity, bit growth is expected to accelerate in the mid to late 2026. SanDisk may further accelerate capacity ramp-up, and the commissioning of the new K2 factory by the joint venture will give it more flexibility in expanding capacity in the second half of 2026. Part.04 Valuation and Target Price Adjustment Logic Can NAND follow the path of DRAM in leveraging AI to drive profitability and multiple expansions? If enterprise solid-state drives can be confirmed as a structural replacement for hard drives, this will significantly strengthen the long-term bullish logic. However, we currently believe that the accelerated growth of enterprise solid-state drives in 2026 will be one to two years beyond the trend in this long-term growth area. Of course, SanDisk also needs to penetrate this market significantly, which will enhance investors' confidence in sustained profitability exceeding historical levels. Can growth momentum spread from a single terminal market to more areas? Although the outlook for the data center market is good, unlike DRAM, NAND still relies heavily on areas such as smartphones and PCs. Therefore, unless the first point above becomes a reality, more participation from the PC and smartphone markets is needed to increase cyclic demand and prolong profit sustainability. The target price has been raised to $230, with an optimistic scenario target price of $300. Considering the 2025 fiscal year, the average EPS for the past 9 years is estimated to be $6.33. If the stock price is $200, the corresponding P/E ratio for this profit level is about 32 times.