Morgan Stanley has returned Nvidia to the top spot on its semiconductor conviction list, arguing that the stock’s recent lack of price appreciation does not mirror what it sees as improving underlying business trends.
Analyst Joseph Moore said in a Monday note that "For the last two quarters NVIDIA has not moved while business has continued to strengthen - a function of concerns about the durability of current growth." He adds that the stock trading at 18x 2027 earnings per share (EPS) presents "a surprisingly good entry point."
That repositioning reverses an earlier stance in which Morgan Stanley favored memory-related names such as SanDisk and Micron on the view that, at the time, AI-driven earnings leverage appeared stronger in those parts of the industry. Moore noted that the memory trade produced far larger returns than expected - memory stocks rose in the range of 300% to 900% since that call - while Nvidia’s share price remained roughly flat even as consensus earnings expectations for the company rose by 38% over a six-month span.
Moore attributes the disconnect between fundamental momentum and stock performance to investor concerns about the durability of AI-driven growth and potential market-share pressures. He expects those worries to ease as Nvidia rolls out upcoming product updates that he believes will reinforce the company’s leadership roadmap and directly address competition-related questions.
On the demand side, Morgan Stanley’s supply-chain checks point to hyperscalers committing to multiyear increases in spending, the analyst wrote. In some instances, customers are reportedly paying upfront to secure capacity through 2028, a behavior Moore interprets as evidence the current investment cycle is ongoing rather than peaking.
Moore also observed that while constraints on AI processor supply could relax in the coming months, loosening GPU shortages would likely play into Nvidia’s hands. He argues that if other system components - memory, storage, optics and power - become the primary bottlenecks, Nvidia may still benefit as system-level constraints rearrange the competitive landscape. The analyst cautioned that investor emphasis on near-term supply indicators can obscure the underlying strength of demand, and he suggested that any reduction in GPU lead times could catalyze market-share reacceleration similar to patterns seen in prior cycles.
Looking to the calendar, Moore sees the upcoming GPU Technology Conference (GTC) as an event that could help rebuild confidence around Nvidia’s market-share outlook, even if it does not fully resolve longer-term questions about the durability of AI capital spending. He expects GTC to underline Nvidia’s broader ecosystem advantages and rack-level strengths; he also noted the potential role of Groq intellectual property in the company’s roadmap.
On competitive positioning, the analyst acknowledges some erosion in moats as several large customers adopt more architecturally agnostic approaches. Even so, ongoing checks reportedly show a continued preference for Nvidia in many deployments. Morgan Stanley’s analysis pegs Nvidia at roughly 85% revenue share in AI processors, with ASICs accounting for just over 10% and AMD at just under 5%.
Moore also highlighted expectations that major ASIC and AMD users will expand their Nvidia business materially, with those customers forecast to grow their Nvidia-related spending by more than 80% in 2026. He noted that while some peers may post slightly faster growth in 2026 - a function of Nvidia’s large scale - Nvidia’s positioning remains strong across multiple vectors.
Following the reassessment, Morgan Stanley kept an Overweight rating and a $260 price target on Nvidia. The bank frames the key investment question as confidence in the durability of AI-driven demand rather than a focus on near-term results. Moore observed that when visibility into future growth improves - as it has in past cycles - Nvidia shares have tended to produce periods of meaningful outperformance.
Overall, the note portrays the recent pullback in Nvidia’s share price as a potential opportunity for investors willing to buy into the thesis that the multi-year investment cycle among hyperscalers is intact and that product and ecosystem advantages will support market-share resilience.