By Steve Garmhausen

U.S. tech-related stocks are still struggling to recover from Monday's selloff, which was sparked by news that Chinese artificial-intelligence company DeepSeek had developed an advanced AI model on the cheap. By basically MacGyvering an AI system for a reported $5 million, DeepSeek challenged an industry based on massive investments and cutting-edge hardware. The S&P 500 ended Monday down 1.5%, while the Nasdaq composite fell 3.1%. Nvidia, a major player in AI chips, led the plunge, with its stock falling by 17%.

So how concerned should investors be, especially those who have overweighted big U.S. tech names? We put that question to several wealth managers and a fund manager in this week's Barron's Advisor Big Q.

Chris Grisanti, chief market strategist, MAI Capital Management: I don't think this is a false alarm. I think it's a real threat to expensive semiconductor chip manufacturers. It will lower the barriers to entry in the AI business, and it's a huge deal. Until now you needed tens of billions of dollars of capital expenditure to build data centers to make these large language models. DeepSeek is saying that you don't need all that to create something that's maybe not cutting-edge but is good enough for 90% or 95% of the applications that AI will be used for in the near term.

There are questions about whether what DeepSeek is claiming is real. Did they actually use the really expensive chips? Did they really do it for five million bucks? And I've concluded that they probably are [being honest]. Everything's open source; you can read their code and see how they're doing it. They revealed in great detail how they engineered it, and that they did a whole bunch of steps that you wouldn't do if you had the really expensive chips. It seems on its face to be a credible competitor.

I've been looking for metaphors for this development, and the best I could come up with was Henry Ford, who didn't invent anything. He didn't invent the automobile. He didn't invent the assembly line. He didn't invent interchangeable parts. But he put them all together, and he made them work really well. And that's kind of what I think is happening here.

From an investor standpoint, I am now suspicious of the high multiples that are given to Nvidia and other companies, for example those that are producing the energy or are going to produce the energy for AI computing. And I don't think that's a temporary thing. This headwind isn't going to go away. I don't think we're going to look back at this and say, "Oh, remember that blip from DeepSeek?" I think this is more serious.

Evan Feagans, portfolio manager, TCW Artificial Intelligence ETF (AIFD): I think Monday's market move was a drastic overreaction. This is potentially a very positive breakthrough on how AI models are architected. It's potentially very good if you think about long-term AI adoption. The base case of how much AI we're going to be using in five years has increased because we can get more performance for a better price and address more use cases.

The reason I think it was a big overreaction in terms of the hardware names like Nvidia and others is that that infrastructure is still needed. It's still a huge component to making these models better. I think Monday's reaction was predicated on the assumption that [the current] level of performance is good enough for the foreseeable future. And I don't think that's true.

From an investing standpoint, I would say that this was a short-term bump. The last time AI had a real hiccup was in August and September. The debate then was about the return on all of this AI investment. Would it continue to scale? Would it continue to get better? Will these companies get a return? Will there actually be good use cases? And it's funny, because now we're getting a breakthrough where it's scaling, and the market is down on it again, even though it's directly hitting against the last bear case.

Jonathan Shenkman, president, chief investment officer, ParkBridge Wealth Management: For the past 10 or 15 years people have concentrated their investments in the S&P 500, and in particular the large tech stocks, and they didn't see any risk associated with it. What happened over the weekend was really one big story: a company implementing something could change the whole dynamics of the market. Nvidia, which was the stock market darling for years, fell about 11% in a few hours. It's a reminder that things can change on a dime. That's why diversification is so important, and hopefully this illustrates the point to people. Unfortunately, people don't change. It's just the way the world works. This will scare people for a little bit, then they'll go back to finding the next hot stock and they'll concentrate their investments there. That's how people are. Human psychology doesn't change, unfortunately.

Robert Levitt, founder and chief investment officer, Levitt Capital Management: This is not a devastating issue for the market. It is for the S&P, because the Magnificent Seven stocks are overweight in the index. But this might just set back these AI stocks a little bit for a short period of time as they cool from their previous levels of growth.

We're not making changes to client portfolios, because we are diversified in the way we develop our portfolios. We have a weighting to technology, but it's much closer to a weighting like the Dow than the S&P. The S&P last year was up 27%, and the Dow was up about 12%, and both were 100% equities. So if our portfolios were somewhere in the middle, with 60% equities, we look like geniuses. And we're not geniuses. It's just that with something like AI, one day everybody loves them and the next day they don't. There's an emotional component to them. But that doesn't change their value and their importance going forward.

This content was created by Barron's, which is operated by Dow Jones & Co. Barron's is published independently from Dow Jones Newswires and The Wall Street Journal.

(END) Dow Jones Newswires

January 29, 2025 15:34 ET (20:34 GMT)

Copyright (c) 2025 Dow Jones & Company, Inc.