Nvidia (NVDA) stock has struggled to keep pace with the broader semiconductor rally this year, but Morgan Stanley just says the AI chip giant remains one of its favorite names.
Shares of Nvidia fell 3.5% on Monday as investors continued to question whether the chipmaker can sustain its growth. Through July 13’s close, Nvidia was up 9.1% year to date, well behind the Philadelphia Semiconductor Index’s 72.2% gain over the same period.
The muted stock performance comes even as Nvidia’s biggest customers continue to ramp up AI spending. On Monday, Meta Platforms (META) said it would increase spending on its Louisiana AI data center to more than $50 billion. Meta, along with companies such as Microsoft (MSFT), Amazon (AMZN), and Alphabet (GOOGL), remains one of Nvidia’s largest buyers of AI chips.
Since generative AI took off in 2023, Nvidia has dominated the AI accelerator market. Its GPUs power large-language-model training and inference workloads, helping Nvidia become the world’s most valuable publicly traded company.
Still, investors have become increasingly cautious. Hyperscale cloud providers are developing their own custom AI chips, while many on Wall Street continue to question whether hundreds of billions of dollars in AI infrastructure spending will ultimately generate attractive returns.
Wall Street, however, remains overwhelmingly bullish. According to TipRanks, 37 analysts covering Nvidia have an average 12-month price target of $309.33, implying roughly 52% upside from recent levels.
Getty Images
Nvidia earnings show AI demand remains strong
Nvidia’s latest results continued to support the AI investment thesis.
For the fiscal first quarter ended April 26, Nvidia reported non-GAAP earnings of $1.87 per share, beating Wall Street’s estimates of $1.76. Revenue came at $81.6 billion, up 85% from a year earlier.
Related: Major AI chip stock plunges after blockbuster $26.5 billion Nasdaq debut
The chipmaker also reported record Data Center revenue of $75.2 billion, up 92% year over year.
“The buildout of AI factories — the largest infrastructure expansion in human history — is accelerating at extraordinary speed,” said Nvidia CEO Jensen Huangin a statement. “Nvidia is uniquely positioned at the center of this transformation as the only platform that runs in every cloud, powers every frontier and open source model, and scales everywhere AI is produced.”
Nvidia’s upcoming Q2 FY2027 earnings report is scheduled to be released in August. The company expects fiscal second-quarter revenue of $91 billion, plus or minus 2%, assuming no data center compute revenue from China. It also forecasts an adjusted gross margin of 75% and adjusted operating expenses of about $8.3 billion.
Morgan Stanley says Nvidia’s growth story is getting broader
After hosting investor meetings with Nvidia’s management, Morgan Stanley reiterated its overweight rating and a $288 price target, according to a recent research note sent to TheStreet. Nvidia remains the firm’s top semiconductor pick.
Morgan Stanley said Nvidia management described “accelerating growth rates” even as revenue approaches “$100 bn per quarter.”
Related: Apple stock move vindicates Palantir CEO warning for AI industry
Rather than relying solely on hyperscale cloud providers, Morgan Stanley said Nvidia is seeing growth from three major customer groups: AI labs, hyperscalers, and enterprise, industrial and sovereign AI customers.
The bank noted that AI labs currently represent about 20% of Nvidia’s demand. It also said Nvidia’s exposure to one leading frontier AI model has increased from minimal levels to “close to 50%”, while other frontier models continue to be built primarily on Nvidia hardware.
Morgan Stanley also dismissed concerns that custom AI chips will significantly erode Nvidia’s market share.
“We continue to believe that two things can be true at once: hyperscalers will develop and deploy custom silicon alternatives (ASICs), while Nvidia will retain a very large portion of the business,” the analysts wrote.
The firm said its industry contacts support management’s argument that “the lowest cost per token is quite frequently from Nvidia,” adding that cheaper custom silicon “does not drive better token economics.”
“With both Nvidia’s and Broadcom’s AI businesses expected to grow more than 80% next year while remaining supply constrained, we don’t see a dramatic shift going forward and remain enthusiastic about growth in both categories,” Morgan Stanley said.
The bank also highlighted accelerating demand from sovereign AI projects, enterprise customers and neocloud providers, saying power constraints, reshoring efforts and geopolitical considerations are creating a new wave of AI infrastructure investment. Management described that opportunity as fragmented today but one that could reach “substantial scale when the time comes.”
Morgan Stanley also said Nvidia pushed back on recent reports that Rubin Ultra could be delayed until 2028, telling investors “Rubin Ultra will ship next year.” While management acknowledged changes to the rack design, it characterized them as improvements rather than delays.
Although Nvidia’s size could limit further multiple expansion, Morgan Stanley said it now views the company as “the best value in the group.”
“The stock has risen but underperformed many peers. However, our conviction remains high. We previously rotated our top pick to Sandisk and later Micron because those names offered greater leverage to parts of the AI supply chain, but we now believe Nvidia offers the best value in the group,” Morgan Stanley said.
Related: Citi sends powerful sign to SpaceX investors
