Chips' "first choice" NVIDIA, replacing storage—Morgan Stanley's reasoning: the stock price has been stagnant for a long time, and the "peak growth in 2026" theory has already been discredited

Nvidia’s stock price has hardly moved in the past two quarters, but Morgan Stanley’s view is that the fundamentals are actually strengthening. The market is just stuck on two issues—whether growth can “sustain” through 2026 and whether market share will be gradually eroded by ASICs and AMD. Morgan Stanley has upgraded Nvidia back to the semiconductor sector’s “Top Pick,” maintaining an overweight rating and a $260 target price, viewing the current valuation as a rare entry point.

According to Chase Wind Trading, Morgan Stanley’s North American semiconductor analyst Joseph Moore stated in a recent report that Nvidia’s 2027 earnings multiple of about 18x is an unexpectedly good entry point. His logic isn’t betting on next quarter exceeding expectations, but on the market’s skepticism about “growth durability” beginning to loosen in the coming months.

Supporting this view are more signals from the supply chain and customer behavior: the report mentions that large cloud providers are locking in longer-term upstream orders, even with some “3-year orders and partial full prepayments.” Such actions are hard to reconcile with plans to “slam on the brakes” next year. However, the report also admits that downstream cash flow pressures are a real issue: some cloud infrastructure businesses are operating with negative cash flow, and financial projections suggest capital expenditures may slow next year.

Regarding market share, Morgan Stanley does not deny that Nvidia might see a slight decline in 2026: the structure they present is about 85% revenue share for Nvidia, slightly over 10% for ASICs, and under 5% for AMD. Given Nvidia’s scale—approaching “$80 billion per quarter”—a slightly faster growth rate among peers and a 1–2 percentage point share retreat is not surprising.

The essence of the two-quarter sideways movement: the market is betting on “growth not being durable,” rather than denying current prosperity

Morgan Stanley attributes Nvidia’s recent “stagnation” to structural expectations: the market has already priced in the current strength, and the real disagreement lies in 2027.

It highlights an unusual comparison—after shifting sector favorites from Nvidia to SanDisk, then to Micron, the memory stocks gained between 300% and 900%, while Nvidia’s stock “remained flat”; meanwhile, Nvidia’s “current quarter” earnings expectations were revised upward by 38% over six months. This indicates that the stock’s stagnation isn’t due to data, but skepticism about the cycle’s longevity.

The report also places this sentiment fluctuation into a historical pattern: over the past three years, each year at the start of the year, the market has been uneasy about “the next year,” until visibility improved, at which point stocks temporarily outperformed. It expects 2026 to repeat this process.

Supply chain signals are more “hard”: three-year lock-in orders, even full prepayments

Morgan Stanley believes that proving sustainability is more difficult, but it sees increasingly extreme upstream behaviors. The report notes that large customers are placing 3-year orders with memory suppliers, in some cases even paying in full upfront; an extreme example is “receiving 100% of 2028 revenue this quarter,” with order volumes still several times current levels.

They view these prepayments as “durability clues”: if customers truly plan to slow down next year, it’s hard to explain why they would pre-lock future supply with cash. The report emphasizes this is just one of many signs—feedback from the supply chain indicates multiple links are preparing for sustained growth.

Downstream counter-evidence comes from cash flow: some cloud infrastructure businesses are still “burning cash”

Morgan Stanley doesn’t dismiss “cash flow issues” lightly. The report mentions that some cloud infrastructure operations are in negative cash flow: this can be sustained (balance sheets are strong, and financing options exist), but from a financial perspective, expenses are likely to “decelerate” next year.

Their explanation is more akin to a “time lag”: Nvidia management responded in calls that cloud vendors will earn more from these investments than the market models suggest.

The report also adds that cloud GPU margins are high, and market participants are discussing a “daily widening of the compute supply-demand gap in single digits.” In other words, short-term cash flow may look poor, but that doesn’t mean returns on investment are absent.

Share might slightly retreat, but customer focus has shifted to Rubin

On the competitive landscape, Morgan Stanley provides a clear breakdown: Nvidia’s revenue share is about 85%, ASICs slightly above 10%, and AMD below 5%. They expect peers to grow slightly faster in 2026, with Nvidia’s share possibly declining by 1–2 percentage points—due to the “physical” reality that larger scale and supply constraints make maintaining the same growth more difficult.

However, the report believes that customers are mainly focused on Rubin (expected to ship in the second half), while prior competition mainly involved Blackwell versus various alternatives. It also admits that the moat has “been somewhat eroded,” as leading model developers prefer architecture neutrality, using Nvidia, AMD, Google Cloud TPU, and in-house ASICs.

But according to their research, even in areas emphasizing “lower total cost of ownership,” customer preferences often still favor Nvidia. The report even mentions that the two largest ASIC users and the two largest potential AMD users in 2026 could still see their business grow by over 80% with Nvidia.

Additionally, for more aggressive bullish narratives—such as frontier model developers requiring “hundreds of GW of compute” by 2029 (versus about 30 GW this year)—Morgan Stanley remains cautious: it doesn’t buy that number but also sees no signs of a “cycle end” by 2026.

Its key constraint is capacity: semiconductor capacity cannot be ramped up to that scale in a short period. More importantly, feedback from the global supply chain indicates not “weaker orders,” but “cash pre-locking growth through 2028.”

For the upcoming GTC conference, Morgan Stanley expects it may not fully convince skeptical investors of the sustainability of capital expenditures, but it will likely clarify the market share debate. This event will resemble 2024—offering a more comprehensive four-year roadmap, emphasizing that competition involves not just chips but also rack and ecosystem development; it also expects Groq IP to feature in the roadmap.

Risk Warning and Disclaimer

Market risks are present; investments should be cautious. This article does not constitute personal investment advice and does not consider individual users’ specific investment goals, financial situations, or needs. Users should consider whether any opinions, viewpoints, or conclusions herein are suitable for their circumstances. Invest at your own risk.