Why market panic over China's DeepSeek is 'overblown,' analysts say

Don't jump off the Nvidia bandwagon just yet.

China’s AI startup DeepSeek triggered a tech sell-off today as investors panicked over fears of a cheaper open-source large language model, raising concerns about the US’s AI dominance.

The tech-heavy Nasdaq ( ^IXIC ) closed down 3%, with Nvidia ( NVDA ) leading the decline. The chip giant’s stock plummeted 16.9%, erasing a record $589 billion from its market cap.

That sent a ripple across the broader technology sector. Chip stocks including Broadcom ( AVGO ), Lam Research ( LRCX ), KLA ( KLAC ), and Marvell ( MRVL ) fell Monday, while Magnificent Seven members Microsoft ( MSFT ) and Alphabet ( GOOGL ) dropped 2% and 4%, respectively.

The fall adds pressure for the group ahead of fourth quarter earnings. Wall Street is already expecting the group’s profit growth to fall to 22% in the fourth quarter, its slowest rate in nearly two years, amid renewed focus on capital expenditures.

Watch: Nvidia rout highlights 'concentration risk' in markets

But the panic-selling may be overdone, according to top analysts. Bernstein’s Stacy Rasgon characterized the market's reaction as "overblown," telling me the development of DeepSeek doesn’t spell out “doomsday for AI infrastructure."

“I am not of the belief that we're anywhere close to the cap on compute needs for artificial intelligence,” Rasgon said. “I'm of the belief that if you're freeing up compute capacity, it likely gets absorbed … We're going to need innovations like this if you're going to be able to keep things going.”

Futurum’s chief strategist Daniel Newman echoed this sentiment . He said a more efficient model like DeepSeek will expand AI use, citing an economics concept called the Jevons Paradox.

“The market is completely missing this one,” Newman warned. “If we can use compute more efficiently … the companies that we're saying aren’t driving enough revenue will be able to build their models cheaper. They'll be able to create solutions with less overhead expense, and they're going to drive more EPS.”

The idea of the Jevons Paradox was also cited by Microsoft CEO Satya Nadella, who tweeted that as AI gets more efficient, its use will skyrocket.

“Spending is still continuing to accelerate … I don't think that this has to stop that,” Rasgon added.

And it's perhaps that sentiment that helped alleviate some of Wall Street’s initial panic as Big Tech names including Microsoft closed well off their lows of the session, while Amazon and Meta finished Monday up 0.3% and 1.9%.