Morgan Stanley Remains Bullish on Nvidia Despite DeepSeek: Are Current Levels Appropriate for Buying?

The recent tech stock selloff, fueled by DeepSeek, has made Nvidia shares highly attractive for purchase, according to Morgan Stanley.

Analyst Joseph Moore once again highlighted Nvidia as a top pick and reiterated his overweight rating on the stock in a note to clients on Thursday. His price target of $152 for the dominant AI chipmaker suggests a potential upside of approximately 22% from Wednesday’s closing price.

“Although sentiment has worsened due to potential long-term risks, short-term business remains strong, visibility on Blackwell chip supply is increasing, and customer willingness to spend is evident,” Moore wrote. “We remain very optimistic about how the rest of the year will play out.”

Nvidia shares have surged nearly 85% over the past year but took a hit following the emergence of Chinese AI startup DeepSeek. The startup is using Nvidia’s less efficient chips to develop an AI model that rivals OpenAI’s ChatGPT at a much lower cost than what U.S. tech giants are investing in the AI race.

DeepSeek purchased 10,000 Nvidia A100 chips—first released in 2020 and two generations behind Nvidia’s current Blackwell chip—before A100 sales to China were restricted at the end of 2023, according to Stanford University’s Cyber Policy Center.

The tech market selloff on January 27 caused Nvidia to plummet by 17%, marking the largest single-day market value loss in history. Nvidia, which is set to release its financial results on February 26, is already down more than 6% this year.

According to Moore, DeepSeek presents some challenges related to export controls and long-term AI investments, but short-term catalysts, such as Nvidia’s Blackwell and Hopper chip solutions, remain intact.

Demand for Blackwell chips remains strong, and “very promising” signals for future demand were noted with CoreWeave’s announcement on Tuesday that they have integrated Nvidia GB200 NVL72 instances into their platform. This makes CoreWeave the first cloud service provider to offer Blackwell at scale, Moore said.

The analyst continues to support Nvidia, emphasizing that capital expenditure comments from the company’s largest customers reinforce their AI investment plans. For example, Nvidia’s cloud clients remain committed to purchasing additional GPUs in their efforts to boost revenue, he added.

“For investments that are not generating revenue today, there remains a strong commitment to advancing modern technology,” Moore said. “Many architects of the largest [General Artificial Intelligence] clusters have reaffirmed their commitment to scaling up large training clusters, with no indications that DeepSeek will alter this dynamic,” the analyst added.

Looking ahead, Moore believes that Nvidia’s biggest long-term catalyst goes beyond AI model training and is centered on the company’s leadership in the inference market, especially as these tasks become increasingly complex. Inference refers to the process where a trained AI model applies its knowledge to new data, making predictions or decisions based on it.

“We remain convinced that Nvidia is the biggest winner from long inference workloads,” he said.