Bonus Episode: What Is DeepSeek and Why Is It Sinking Stocks?
08:51 Share

Capital Ideas, conversations with Mike Gitlin from Capital Group, features our top investment professionals sharing what drives them in today's market. Get stories and actionable insights. Subscribe wherever you get your podcasts. Capital Client Group, Inc. Bloomberg Audio Studios. Podcasts. Radio. News.

Good morning, I'm Nathan Hager. And I'm Karen Moscow. Here are the stories we're following today. And Karen, we begin with concerns over the impact that a Chinese artificial intelligence startup could have on big U.S. tech stocks. It is called DeepSeek. Buzz grew over the weekend about DeepSeek's latest AI model being cost-effective while running on less advanced chips. That's casting doubt on the validity of the sky-high valuations for companies like Intel,

NVIDIA. We get more from Bloomberg Markets reporter Mark Cudmore. I think this is probably a game-changing moment because I think suddenly the whole U.S. stock exceptionalism of the last 14, 15 years is probably undermined by this because it simultaneously is good for productivity and growth in the rest of the world, good for companies all over the rest of the world, and undermines the whole U.S. tech phenomenon. So I think it hits both sides. So I think actually last Thursday we probably saw the long-term peak in U.S. share of global market cap.

Bloomberg's Mark Cudmore says the AI model from DeepSeek is widely seen as competitive with OpenAI and Meta's latest offerings. Nathan, Chinese startup DeepSeek is now top of the iPhone download charts despite being hobbled by sanctions. Capital.com senior market analyst Kyle Rada says it raises questions about the market narrative around U.S. tech dominance.

If anyone's on a sort of group chat with equity investors or any kind of equity strategist, phones are running hot effectively talking about, well, what is this product? How have they achieved these outcomes? And what does this mean for the AI investment boom in the United States? You know, it raises concerns about overinvestment in artificial intelligence. It raises concerns that perhaps the United States might have the competitive advantage that it has over the rest of the world. It raises concerns that chip demand won't be as great going forward either, and companies like NVIDIA are overvalued.

Capital.com Senior Market Analyst Kyle Rotta. Several of the big tech stocks falling on the news. Nvidia shares down 10 percent. ASML is off by about 11 percent. Meta and Microsoft are lower by well more than 4 percent. Palantir down 6 percent. Supermicro and SoundHound are off by at least 9 percent. And Karen, this deep-seek news comes just as investors brace for a slew of earnings from the Magnificent Seven. Bloomberg's John Tucker is following this end of the story. And John, an

Feels like the news couldn't get any more dicey. Well, it could. The company's sharers have been near record highs. Their valuations are stretched, Nathan. And to this, the fact that the group's profit growth, it's projected to come in at the slowest pace in almost two years. Profits for the seven giants are projected to increase by...

22% in the fourth quarter from a year earlier. That actually would be the smallest jump since the first quarter of 2023. The seven companies' announcements, they start Wednesday. That's when Microsoft, Meta, and Tesla are scheduled to report. Apple follows Thursday while we get...

Alphabet and Amazon next week, and then chipmaker NVIDIA. That comes late February. CapEx also going to be a big theme. Microsoft, Alphabet, Amazon admit it. They're all projected to have spent more than $200 billion combined on capital expenditures in their last fiscal year. And they've all pledged to spend more in the current year. In New York, I'm John Tucker, Bloomberg Radio.

Coast to coast on Bloomberg Radio. Nationwide on Sirius XM. And around the world on Bloomberg.com and the Bloomberg Business App. This is Bloomberg Daybreak. Good morning. I'm Nathan Hager. We are seeing a rout in tech stocks this morning. Chinese startup DeepSeek is driving a lot of buzz around the idea of a low-cost artificial intelligence model running on less advanced chips.

Capital.com senior market analyst Kyle Rada says it could mark a narrative shift for the U.S.-led AI boom. That kind of proverbial arms race in technology investment and AI investment has taken a very, very interesting turn when now China looks like a very viable competitor, uh,

and a competitor that perhaps might catch the ire of the Trump administration. It's Kyle Radha of Capital.com. And for more, we're joined by Bloomberg Radio and Television News anchor Kriti Gupta. Kriti, good morning. What is it about DeepSeek's model that we're hearing about right now that's got tech investors clearly so spooked this morning?

Well, good morning, Nathan. This is kind of the worst nightmare for a tech bull in that we know that so much money has been pumped specifically from the hyperscalers like Amazon, Microsoft, Apple, etc., who are sitting on trillions of dollars of cash just getting pumped into the future of artificial intelligence.

And that's, of course, what's created this massive stock market rally. The concern here now is that maybe all of that money didn't necessarily need to be put in to the extent that it has in the stock market. Let me walk you through some of my math here. And Nathan, for example, one NVIDIA Blackwell chip is expected to go on the market for somewhere like $30,000 to $40,000 per unit. Remember, we're talking about a chip alone and, of course, a ton, basically, of units per

used for one individual AI data centers, as so many of these companies are trying to build out, of course, hundreds of data centers around the world. So you can see very quickly how the argument of just how much one chip is valued can kind of skyrocket into almost an unknown positive figure. And that's the sentiment, at least, that's been fueling the stock market rally. But

But then lo and behold, DeepSeek, this Chinese startup around AI that's using technology very similar to that of OpenAI, which is already kind of on the front of the latest in AI.

They are now saying that they can do it on a much cheaper cost. So if you're talking about the things you need to get to artificial intelligence, think of it as a recipe. You need chips. You need engineers. You need software kind of talent who are thinking about this. You need a data center. You need an electricity grid. You need things like natural gas or oil to power that data center, even solar, for example, for the folks that are looking into that. Imagine if that

one element of that recipe is now able to be significantly cheaper. That's what DeepSeek has basically shown over the weekend and in going into last week as well, that they can do that. They can provide the same sort of technology for a much cheaper price. So the reason you're seeing the market reflected is saying, well, hold on a second. If you can do this for not as expensive

expensive of a price as previously thought, then why have these stocks been rallying to such an extent? And that's kind of the sentiment that's driving this global tech route today. And it's a sentiment that had driven so much of the rally, as you mentioned, up to now. That puts a focus, a pretty bright spotlight now, doesn't it, on the tech earnings that are coming up as soon as Wednesday for a lot of these big hyperscalers?

It absolutely does. And I think context is everything here, Nathan, because remember these hyperscalers, I mean, we talk about just a handful of companies that are powering NVIDIA. Remember NVIDIA over about 50% of their kind of gains and their customer base is really just from a handful of the mag seven names, your Amazons, your Microsofts, et cetera. And they're able to invest that much because they are still sitting on trillions of dollars of cash post pandemic and

And things that haven't really been able to be deployed in acquisitions. And this is something that the tech playbook is known very well for that, you know, you kind of bolt on some of these deals. You buy new technology. Think of Facebook, for example, buying Instagram back in the day. It's a similar playbook. They haven't really been able to do that and use that cash.

Instead, they're spending a lot of that cash in investment, in research, and of course, in AI. Now, basically, the element here is that maybe you don't need to do that. But I'm going to caveat all of this, Nathan, by simply saying this is still experimental technology. This is still just one example. And we also don't have the full figures on what kind of chips they use, just how much it's costing, and whether or not there are others like it, or if this is just kind of a one-hit wonder. Those are still the questions that remain in this market.

This is Bloomberg Daybreak, your morning podcast on the stories making news from Wall Street to Washington and beyond. Look for us on your podcast feed by 6 a.m. Eastern each morning on Apple, Spotify, or anywhere else you listen. You can also listen live each morning starting at 5 a.m. Wall Street time on Bloomberg 1130 in New York, Bloomberg 99.1 in Washington, Bloomberg 92.9 in Boston, and nationwide on Sirius XM Channel 121.

Plus, listen coast to coast on the Bloomberg Business app now with Apple CarPlay and Android Auto interfaces. And don't forget to subscribe to Bloomberg News Now. It's the latest news whenever you want it, in five minutes or less. Search Bloomberg News Now on your favorite podcast platform to stay informed all day long. I'm Karen Moscow. And I'm Nathan Hager. Join us again tomorrow morning for all the news you need to start your day right here on Bloomberg Daybreak.

In 15th century Florence, the great inventor Leonardo da Vinci dreamt of creating a flying machine. But something kept getting in his way. Admin. Piles of it. Luckily, Leo used the smart buying tools on Amazon Business, so he could work more efficiently. With the extra time, he not only invented the flying machine, but actually built it. Magnifico. Incredibile. Splendido. Whoa, easy there, Leo.

Amazon Business, your partner for smart business buying.