Why Morgan Stanley Raised SIMO So Aggressively
Morgan Stanley’s latest note makes a strong call on the storage supply chain, with the most notable move being its target price revision for Silicon Motion (SIMO.O). The bank lifted its target from $155 to $400, based on 23x estimated 2027 EPS, and said it expects the company’s 2026 revenue to reach a record high. It also raised Longsys from RMB 300 to RMB 673 and Phison from TWD 2,248 to TWD 2,588.
That said, this is not a blanket bullish view on every NAND-related name. Morgan Stanley kept Equal Weight ratings on both Longsys and Phison, signaling that the opportunity is selective rather than universal. The key shift in its framework is that NAND demand is no longer being driven mainly by the familiar smartphone, PC, and consumer SSD inventory cycle. Instead, the bank sees AI data centers pushing the market into a new phase driven by enterprise SSDs, AI boot drives, and long-duration procurement by major cloud providers.
For investors, this changes the lens through which NAND pricing should be viewed. The thesis is no longer simply “SSD prices may rise.” It is that AI infrastructure is altering both the composition of demand and the timing of purchases, which in turn could make the next cycle more dependent on enterprise deployment and cloud capex than on a rebound in consumer electronics.
AI Servers Are Rewiring the NAND Demand Model
Morgan Stanley argues that NAND has historically been pulled around by handset builds, PC demand, and swings in consumer SSD inventory. What is changing now is the rising role of AI servers. These systems do not just consume GPUs and HBM; they also require meaningful amounts of local storage, enterprise SSDs, and boot drives used for system initialization, management, and operational continuity.
Once hyperscalers and core cloud service providers begin sourcing these components through more stable and longer-term agreements, NAND pricing behavior may also change. In other words, the market could become less exposed to short-lived retail and consumer fluctuations and more sensitive to data center procurement, system architecture choices, and enterprise-grade storage availability.
Morgan Stanley’s supply-demand model reflects this shift clearly. It projects total global NAND demand of 1111 EB in 2025, 1250 EB in 2026, and 1484 EB in 2027, against supply of 1128 EB, 1058 EB, and 1347 EB, respectively. That implies roughly a 2% surplus in 2025, followed by a 15% shortage in 2026 and a 9% shortage in 2027.
Importantly, this shortage case is not built on an aggressive consumer recovery. The report explicitly assumes relatively restrained conditions for smartphones and PCs, with per-device NAND capacity broadly flat and end-unit shipments still tracking declines under the hardware team’s model. The shortage outlook instead rests on continued expansion in AI servers, cloud spending, enterprise storage, and boot-drive deployment.
By 2027, Morgan Stanley expects AI-related NAND demand to rise 60% year over year to 609EB, representing 41% of total NAND demand. With total demand seen at 1484EB and supply at just 1347EB in the same year, AI becomes the dominant marginal buyer. The more AI contributes to overall demand, the more NAND pricing and availability become tied to hyperscaler purchasing behavior, server configurations, and enterprise SSD supply constraints.
Pricing Is Already Diverging Across Product Categories
Channel checks suggest that this demand shift is already showing up in prices. In Morgan Stanley’s 3Q26 checks, TLC enterprise SSD pricing was up around 30% quarter over quarter. Server DRAM pricing was up roughly 20%, while legacy DRAM products such as DDR3 and DDR4 rose by around 30% to 40%.
Consumer-facing NAND products, however, showed much smaller increases. The reason is straightforward: smartphone and PC customers face tighter margin pressure and have less capacity to absorb the same magnitude of cost inflation. As a result, the market is not seeing a uniform repricing across all categories of NAND-related products.
This distinction matters. The strongest pricing power is appearing in data-center-linked products, not across every storage segment. For equity investors and industry participants, that means the greatest benefit is likely to accrue to companies with meaningful enterprise SSD, controller, and AI server exposure, rather than to businesses still dominated by consumer channel demand.
Why SIMO Stands Out in This Cycle
The bullish case for SIMO centers on two business lines that align closely with AI storage growth: enterprise SSD controllers and AI boot drive modules. Morgan Stanley sees the company’s MonTitan enterprise SSD business as one of its most important growth engines over the next several years.
Specifically, the bank expects MonTitan to contribute 5%, 13%, and 19% of SIMO revenue in 2026, 2027, and 2028, respectively. At the same time, boot drive modules are expected to begin scaling meaningfully, contributing about 15% of revenue in 2026 and 21% in 2027.
Boot drives are not the headline component in an AI server, but they are operationally essential. They support system startup, administrative functions, and ongoing runtime management. As AI server shipments increase, demand for the related controllers and modules should rise in tandem. That is a major reason SIMO’s market positioning is changing in investors’ eyes: it is no longer being viewed only as a consumer controller company, but increasingly as a beneficiary of enterprise and AI-oriented storage growth.
Still, Morgan Stanley’s valuation framework depends on execution rather than on profits already realized. The $400 target price, tied to 23x projected 2027 EPS, assumes successful ramp-up in enterprise SSD and boot drive products, continued customer onboarding, and no material slowdown in AI server demand. If any of those assumptions fail to hold, the sustainability of that valuation re-rating could come under pressure.
Why Longsys and Phison Were Not Upgraded More Aggressively
Longsys and Phison also stand to benefit from firmer storage pricing and stronger AI server demand, but Morgan Stanley stopped short of assigning them more constructive ratings. The reason lies in the practical constraints faced by module makers when NAND supply tightens.
In a constrained market, original manufacturers are more likely to prioritize output for major cloud operators and core CSP customers. That means the incremental supply available to module companies may not be large enough to unlock the same degree of revenue elasticity. In other words, a rising price environment can help average selling prices and inventory values, but volume access may still be capped by upstream allocation decisions.
This is why target prices can move higher while ratings remain at Equal Weight. Improved pricing supports ASP and inventory economics, and a better enterprise product mix can help margins. But if shipment volume is effectively locked up by suppliers and top-tier customers, module makers may see limited upside in realized revenue growth despite a favorable headline cycle.
The dynamic can be summarized simply: this may be a market where prices rise, but not every participant gets the volume needed to fully monetize that rise.
LTA, TCM, and the Real Turning Point in 2028
Another important thread in the report is the role of long-term agreements, or LTA. These contracts can provide some downside protection against falling prices. Morgan Stanley estimates that Kioxia’s LTA coverage could exceed 50% by 2027. But LTAs are not an unqualified positive.
Micron has also indicated that such agreements often include both price floors and price ceilings. That means they can reduce the risk of severe price collapses, but they can also limit upside when shortages become extreme. In a highly constrained market, suppliers may not capture the full benefit of spot-like price spikes if those ceilings are binding.
Module makers, meanwhile, hope to use models such as TCM to shift more inventory burden to customers and stabilize long-term gross margins in the 25% to 35% range. But even that depends on customer acceptance, supply tightness, and whether the products involved are differentiated enough to justify the structure.
The biggest boundary to the current bullish thesis lies in 2028. Under Morgan Stanley’s base case, even in 2028 the market could still face roughly a 5% shortage if AI NAND demand continues to grow 60% year over year and YMTC capacity stays around 310kwpm. However, if YMTC expands capacity to 470kwpm and AI growth slows, the NAND market could shift from shortage toward balance or even oversupply.
This is why the long-term question is not simply whether SSD prices can keep rising. The more important issue is whether AI demand will remain strong enough to absorb new supply over the next two years. For companies like SIMO and other participants tied to AI storage and controller demand, 2026 may mark the beginning of a meaningful enterprise and AI ramp. But for the broader NAND cycle, the decisive variables are likely to be YMTC’s expansion pace in 2028, the strength of CSP capex, and whether suppliers maintain discipline on the supply side.