Motley Fool Money - How Banned Chips Get to China

Episode Date: August 28, 2024

The U.S. has tried to prevent advanced microchips from getting into the hands of its military rivals. They’ve only had partial success. (00:21) Alicia Alfiere and Mary Long discuss results from retailers, Abercrombie’s turnaround, and Chewy’s commitment to customers. Then, (12:40) New York Times reporter Ana Swanson talks with Ricky Mulvey about how Nvidia chips are ending up in China, despite America’s efforts to keep them stateside. Read Ana’s story here. Companies discussed: CMG, JWN, ANF, CHWY, NVDA "With Smugglers and Front Companies, China is Skirting American A.I. Bans." https://www.nytimes.com/2024/08/04/technology/china-ai-microchips.html Host: Mary Long Guests: Alicia Alfiere, Ana Swanson, Ricky Mulvey Engineer: Tim Sparks Learn more about your ad choices. Visit megaphone.fm/adchoices

This episode is brought to you by Indeed. Stop waiting around for the perfect candidate. Instead, use Indeed sponsored jobs to find the right people with the right skills fast. It's a simple way to make sure your listing is the first candidate C. According to Indeed data, sponsor jobs have four times more applicants than non-sponsored jobs. So go build your dream team today with Indeed. Get a $75 sponsor job credit at Indeed.com slash podcast. Terms and conditions apply.

We've got a look at retail and how advanced AI chips are getting smuggled into China. You're listening to Motley Full Money. I'm Mary Long, joined today by Alicia Alfieri, Alicia. Thanks for being here on Motley Full Money. So glad to be here. So the big news this week is that Nvidia is reporting, but that doesn't happen until this afternoon. So this morning, we are going to turn our attention to a completely separate sector of the economy and talk retail. I feel like throughout the summer we've been hearing two different stories about

the consumer in the state of them. On the one hand, retail spending in July was higher than expected. So kind of pointing to signs of a healthy consumer. But on the other hand, we've heard stories from Starbucks and McDonald's and a number of other companies that consumers are tightening their wallets and really feeling high interest rates and are cracking down on discretionary spending as a result. We've got a number of retailers reporting this week. What are you keeping an eye on as these companies come out? Yeah, I've been looking at several companies. And it's been really fun to take a look at how different companies have been faring in this period, right? We have some food chains that have been struggling while others have been doing well. Chipotle, for example, continued to execute

its revenues were up, number of stores continued to expand, it continues to generate cash, and the company continues to train, track, and execute on improved throughput. So that sort of thing is always fun to see and to study. There are so many different types of consumers, and I think that that really hits home when we focus in on results from specific companies because they each cater to a different demographic. Nordstrom reported yesterday and they've got an eye into different types of consumers with their store. You've got Nordstrom Rack and the department store. Did they come out with any different trends that kind of signify or suggest different strengths within those two types of consumers? Well, I would say it's, I would be cautious to use one company's earnings

report as a measure of consumers or economic health. Right. individual companies have different strengths and weaknesses, competitive advantages and management styles that can shape unique performances. I think that's why it's important to really take a sampling across different companies to see what's happening before you explain a company's performance based on macro factors or that sort of thing. One retailer that reported this morning was Abercrombie and Fitch, and they boasted a 127% increase in their earnings. It's sixth consecutive quarter of triple-digit earnings growth. The company also notched its first quarter of over a billion dollars in revenue. Operating income nearly double this time from last year. That all sounds

really good to me. And yet last time I checked this morning, shares were down nearly 18%. What is Wall Street seeing there that I am missing? Yeah. So first, let's dive a little bit deeper into the results. So in the second quarter, net sales were $1.1 billion, which is a record for the company, and up 21% year-over-year. They saw growth across each of its brands, which saw their highest sales in brand history and across all-world regions that they operate in. Gross profit for the quarter was also up, and the company raised its full-year guidance, despite a somewhat uncertain economic backdrop, as the CEO had said.

So the company is now guiding for 12 to 13% year-per-year growth in sales. That's up from a forecast of 10% growth just last quarter, quarter for the full year. And the company is also guiding for 14 to 15% operating margins this year, which is better than its guidance of 14% margins in a previous quarter. But it's that guidance that the market probably had a problem with. Because even though that margin guidance was raised, it's below a 15.5% margin that the company reported this quarter, which could have the market worried that margin growth could be under some pressure later this year. And while that's possible, it seems at odds with the company's actual results.

But that can be the danger when a company has a super cheery investor vibe around it. It's that any kind of expectation that's baked in, you know, if those aren't necessarily met, you can have some volatility. Part of that expectation might come from the fact that in the past year, Abercrombie has outperformed Nvidia, which is surprising, especially when you consider all the hype that's around the chipmaker. Just to put some numbers behind that, the retailer stock is up over 230% in the past year. Meanwhile, in videos up about 179%. So that performance is due in large part to the turnaround that's been executed by Abercrombie CEO, Fran Horowitz. What happened to make that turnaround a possibility? Yeah, well, turnarounds take a little while, and this one did take some time.

I think there were some surprising things that happened. First, Abercrombie had a bit of a reputation for being experienced. exclusive and not in the good way. Also, back in the day, their catalogs seemed to focus more on their models' bodies than the actual clothes. And the company worked hard to turn this image around and make their stores and products more inviting. I've also been surprised with the company and how they've targeted different consumers. So they do still have teenage clothes, but they're moving away from that. And instead targeting people in their 20s and 30s. And you could see proof of that because the company recently added a weddings category earlier this year. So that's

been really interesting to watch. It's one thing to like sit where we are now and to look back and celebrate Abercrombie success in hindsight after it's already happened or while it's in the process of happening. Currently, the company's trading at about 14 times earnings. Is that a fair price to pay for a company that's got this good of a track record? Is more to come? More of that good to come? Or is this kind of the end of the story for them? It's a good question. 14 times forward earnings isn't exactly cheap, and it's important to highlight again just how much the share price has increased over the last year, right? Over 230 percent. And remember, a company that has a cherry investor consensus can be more prone to volatility as expectations can have a cost.

I also like to look at price to simple free cash flow, which is roughly 15.5 times, and that's not too bad. I think what's really important is what do you think the company can do from here? If you were betting on a turnaround a year or so ago, is it fair to say that the company has turned around? If you think so, now you have to look at where the company goes from here, what your thesis is now and how that factors all into the price. Another retailer that reported this morning saw far slower growth than Abercrombie reported, but it's eliciting the opposite market reaction. Chewy stock is up about 15% on the news of the company's becoming more efficient. What else stood out to you in this report? Yeah.

So what really stands out to me for Chui is its deep customer relationships and how they're growing. That's what drove Chui's net sale increases. It was about 2.6 year over year. And it is, again, vital to its growth. So let's kind of look a little deeper into Chui. So first, for Autoship. Their Autoship program is free to sign up and it gives customers automatic deliveries on products that they use all the time and access to additional savings to.

In the second quarter, auto ship sales were over 78% of net sales for the company that's up from 76.1% in last year's quarter. And this is an interesting idea because it's kind of similar to, or it's the closest a retailer could get to those recurring revenues that software as a service companies would expect, right? Not a pet retailer. Another important indicator that I really like to look at for Chui is net sales per active customer. In the second quarter, customers spent an average of $565 on Chewy, and that's up 6.2% year-over-year. And Chewy's health care and specialty business really helped to grow that. Now, this is an important number to watch going forward, as Chui's customers tend to spend more the longer that they're customers with Chui, which shows just how much those deep customer relationships can grow over time. And Chui has previously said that they think they can increase this number.

in the long term to something like $1,700. And that is a pretty big goal. But they're hoping that they can get there with more than just pet food and toys, things like their vet clinics, which they're still in the early stages of building out, could really help them reach that goal as well. But again, that is a long-term goal. You talk about deepening customer relationships.

How important is it, if that's the focus, how important is it for Chui to widen their customers? And what I mean by that is that customer, account increased sequentially, but it did not increase year over year. They've kind of seen this small decline since 2021 when customer account was hit its peak at 20.7 million active customers. Today, they've got 20 million. If, if Chui really can grow the average spend per customer per active customer to that massive, objectively massive number of $1,700 that you mentioned, do they have to worry about customer count as much? Yeah, that is that that is a good question. So I would say

I always thought this was a hot take of mine, but you've already picked up on it, right? The customer numbers are important, right? You don't want to see too much of increasing declines here, right? But the more important thing, I think, for Chui going forward is the deep customer relationships, is increasing that net average sales per customer number. And we also have to remember, you know, Chui saw a lot of customer growth during the pandemic. I think 43% year-per-year-year in 2020 alone. And that's big.

And a lot of people got new pets at the time. I'd argue that if you wanted a pet, you probably got one during the pandemic. But that's a bit of a pull forward, right? So over time, it should get back to normal. And hopefully we won't see as many declines. But to me, the more important number is that net average sales per customer and not quite so much the customer number. And that peak, an active customer count, contributed to the first. fact that the company hit its stock hit its highs in 2021. So today, the stock is down almost 80%

from those highs. What would you say to somebody who bought in at peak chewy time and still believes in the story there, but has seen their stock price go down quite a bit from that high? Yeah. I would say part of our journey as investors is that you'll likely buy shares of a company and at some point those shares will decline. And it can be painful when it falls a lot. It's a risk that many of us encountered during the pandemic. And it can definitely happen as people get more excited about a particular stock or even an industry. This is why it's always important to know why you're investing in a particular company,

know your risk tolerance, and employ a strategy like buying in thirds. That doesn't mean that you stop at three. There are some people who have bought at thirds and are on their twice. a third, right? Or you could take a starter position in a company and add over time as you gain conviction. And I would say also remember what David Gardner often says, which is some stocks will take an elevator down and the stairs back up. So holding good businesses over the long term can help to not only lower your losses, but see the benefits of compounding gains as well. And what say you? Is Chui a good business at the end of the day? I certainly like it.

And I am excited to see how they could continue to develop those customer relationships going forward. Alicia Alpieri, thanks so much for the insight into a couple companies this morning and for closing this out with some wonderful, wise mindset advice. Thanks. You're not supposed to find Nvidia's most advanced microchips for sale in China. The U.S. government banned the export of them. Not everyone plays by the rules, though. Anna Swanson is a reporter for the New York Times covering trade and international economics.

She joined my colleague, Ricky Mulvey, to discuss her investigation on microchips and how they find their way around export vans. Anna, your reporting found that these highly advanced AI chips from companies like Nvidia, Intel, AMD, are being sold in these open markets in China where these advanced chips are not supposed to be sold. To set the table, what are these markets like? What was it like being there? Yeah, it was a really fascinating experience. and kind of really fun and interesting to get to talk to people on the ground there. So I went to one market in Shenzhen that is kind of notorious for being able to get pretty much any electronic imaginable. Shenzhen is right across the border in southern China from Hong Kong. And it's also a huge

electronics capital of China. And so this market stretches for about, half a mile. It's just, you know, incredibly densely packed with vendors selling all kinds of things, but in particular, a lot of computer chips. And so when I went there, I found that I was able to locate some vendors who were offering these chips that have been banned or restricted by the U.S. government pretty easily. We found both a directory online that pointed us to some physical stores. And then I also realized walking around that vendors had on display banned Nvidia gaming chips. And if you ask those people, if you could get these artificial intelligence chips, many times they would say, yes, I can do that for you. I can order that. And, you know,

some people were talking about smaller volumes. But we also ran into some vendors that said that they were moving larger volumes as well, which is pretty significant, something that had not been previously reported at the time. Yeah, one buy for more than $100 million, I believe. I mean, on a personal level, that's incredibly courageous to go there to see if these trade bans are being enforced and to investigate these buys in China. Were you ever nervous being there? Well, I mean, reporting conditions in China are, you know, far from ideal right now. We are, you know, really restricted in terms of the numbers of reporters that are in the country and when we can go into China. But, you know, on occasion, we do have some access and it's really

important to be on the ground there. And you can just learn so much more by being somewhere in person than you can, you know, trying to report about it virtually. So I was really, you know, grateful for the opportunity to speak to kind of real people in China and just get in touch with, you know, what's actually going on on the ground there. One of the things you found on the ground was basically that the price of these servers with highly advanced AI chips, one server with H100 units from Nvidia, goes for about $380,000, which you found is relatively high for an international sale, but there's not like a large black market premium for these chips. Why don't you think there's more of a like a premium? What's that say about the supply chain, you think? Yeah, it's really interesting. I mean, unfortunately, I just have kind of a few data points from my reporting. I, you know, I wish that we had more. But I think what it suggests is that there is, you know, more of a steady supply of these products. Certainly then the U.S. government has recognized previously. So, you know, initially it seemed like after the bands went into place, you know, prices really spiked. But it seems like, you know, from some people, I talked to that they've come down somewhat. And so it may be that, you know, these black markets have found kind of more steady supply chains and found some sources of supply. I spoke with some international

experts about what those prices should be. And it seemed, you know, clear that, like you said, there's somewhat on the high end, but they're still within sort of international, you know, international price ranges, right? And so, you know, again, I don't have a ton. of data. I just have, you know, I did do pretty extensive reporting there, but it's nothing like, you know, extensive market surveys or something. But it's certainly not a picture of scarcity, I guess, I would say that I found in China. And that was pretty surprising to me as well. And to set the table a little bit, what's the spirit of the ban? What was the U.S. government trying to do by banning these advanced, like some advanced AI chips, but not other

advanced AI chips? Yeah. So initially, the U.S. government, you know, said it really just wanted to focus on the most advanced AI chips. And the reason was that these AI chips, while they're useful for a huge range of commercial products, they're also really essential for the most advanced military technologies. So the administration said, you know, their purpose is not to go after the commercial industry, but they kind of became collateral damage along the way because these AI chips are also important for military functions. So the These broader bands on China really started in October of 2022, and they've kind of escalated from then. The U.S. government issued another one to tighten it last year, and then they're also considering further restrictions right now that would apply more to equipment and high bandwidth memory chips, but also ratchet up the tech restrictions on China.

We know the use cases in the U.S. were familiar with self-driving cars and, gaming and that kind of thing, but we often don't think about what the PRC is trying to do with these highly advanced AI chips. What are they trying to do with? What is the government of China trying to do with it? Right. Well, of course, there's also a lot of interest in China in those kinds of commercial aspects. And, you know, I do want to say the vast majority of these chips have been used for things like, you know, algorithms and your social media sites and, you know, very harmless and even helpful things, weather forecasting scientific research. But there certainly is a concerted effort by the Chinese government to also use the most advanced technology to modernize

its military and, you know, to defend Chinese territory and be able to, you know, meet any challenges that it sees to its national security, its national sovereignty. And so, you know, I did, find direct evidence that these types of chips had been used in research in China that related directly to the military, whether it was creating simulations of underwater explosions, like from torpedoes or missiles or modeling what happens when a nuclear weapon detonates. There's a lot of military applications for this kind of advanced technology, you know, most of which relates to the ability to just model and invent things much faster with this kind of advanced computing. And then there's also a suite of concerns around what China could do with AI and quantum computing,

which are maybe a little bit more perspective and sort of in the future rather than actual. But there's a lot of concern about, for example, disinformation or cybersecurity tax and how AI could fuel that as well. We know how hard it is for a government entity, which you're reporting covers the Bureau of Industry and Security, a small government bureau trying to keep track of where these chips are going and stop it. But, you know, how does Nvidia not know where these servers are? They're one of the most advanced companies on Earth. They're building a digital twin of the planet Earth. They're not selling sneakers. And it's not just Nvidia.

It's also AMD and it seems to be Intel. How do these chip makers not know where these servers are? Yeah, it's a great question. I mean, I think like sort of the industry viewpoint is that, you know, they do have practices in place. They say that they all follow the rules. They all carry out due diligence. They ask their customers to comply with export controls. But after, you know, they can monitor the chip going to their customers. But after it leaves their customers' hands, they kind of, you know, lose track of it or, you know, aren't able to track it anymore. There are definitely critics. who say, you know, that's not good enough and also that other industries, you know, have done a better job at tracking products and that the chip industry, you know, just hasn't been pushed by the government to develop the same kind of diligence that, you know, companies that make things that go into nuclear weapons have had to do for a long time. I mean, I think, you know, with certain kinds of chips, there is a good argument that there are already tons of them out there on international markets. They're tiny. They're hard to control. But the most advanced GPUs are different. They're bigger.

They're really expensive. They're really in demand. Right. And they're incredibly profitable for these companies. So I think you can make an argument that, you know, maybe these companies and the government need to think more seriously about the kind of tracking that they're doing. I think the reporting, you know, does kind of suggest that, even though, you know, companies have their arguments about why that is.

is very difficult to do. And then as you're looking at sort of this tech race for the most advanced AI chips, the other company we don't talk about a ton on the show is Huawei. And they're trying to compete with InVidio to build the most advanced AI chips. How are you seeing their technology stack up as you talk to these folks who are in the industry? Yeah, that's a great question. It's a really interesting area. And it's actually, it's pretty hard to tell, in part because,

the Chinese government and, you know, China in general has a real interest in kind of like publicizing and exaggerating the amount of progress that they're making. And the U.S. government has an interest in kind of downplaying that and saying, you know, our controls are extremely effective and they're not being able to, you know, produce these advanced chips at scale. I mean, I think the truth is really somewhere in the middle and we're kind of diving this into this issue exactly with further reporting right now. But it seems like, you know, the controls really have held back China's production of AI chips. But at the same time, they're still making impressive progress. There's been reports recently of Huawei rolling out a new AI chip and Nvidia has recognized

Huawei in its filings recently as a competitor. You know, we see these Huawei chips that there are orders now for them by domestic Chinese companies and they're being deployed in data centers. There's still a lot of uncertainty about exactly how many chips that they can roll out. And certainly, it's far, far less than, you know, people in China would like to have in order to truly develop AI at the pace they want. But yeah, I mean, it does seem like, you know, this program is simultaneously holding China back and then also really encouraging their own efforts to indigenous this technology. And it's still kind of an open question for me, which of those in the end will win out, you know, whether the program will will hamper China more or encourage it to

develop its own AI chips more quickly. Honest Swanson's reporting can be found in the New York Times. The investigation we're talking about today is titled with smugglers and front companies, China is skirting American AI bands. She worked on with Claire Fu. Thank you so much for joining us on Motley Full Money. I really appreciate your reporting, your time and your insight. Thank you. As always, people on the program may have interest in the stocks they talk about. And The Motley Fool may have formal recommendations for or against, so don't buy ourselves stocks based solely on what you hear. I'm Mary Long. Thanks for listening. We'll see you tomorrow.