Good morning,
Thursday was a crazy day in the stock market, with the entire AI trade selling off.
But as Warren Buffett says, "Be greedy when others are fearful."
This market sell-off could be a great opportunity if you're looking at the AI hardware side of the trade, anything going into AI data centers.
The AI buildout is a long-term thing. These projects are in active development, and companies aren't going to shift on a dime.
Today, I'm going through my watchlist and ranking five companies that are selling off but that I think are still good buys long term.
👉 Add these companies to your free watchlist to follow them through the sell-off.
1. Nvidia: Earnings have outgrown the stock
📊 Research NVDA: Nvidia ($NVDA)
It's crazy to say a stock that has done so well has lagged its business. But over the past five years, Nvidia's earnings per share and free cash flow have gone up far more than its share price.
This is a free cash flow machine. It doesn't have to build the factories that manufacture its chips. TSMC ($TSM) does that, which leaves Nvidia with a much more capital-light business.
My Plan: I'm building my position. If Nvidia keeps growing earnings as expected, it doesn't even need a higher P/E multiple to look attractive. In a risk-adjusted sense, it's my first choice on this list.
I was buying more Nvidia ($NVDA) when I filmed.
2. Broadcom: Google's chips are part of the opportunity
📊 Research AVGO: Broadcom ($AVGO)
When I think about Broadcom, I'm mainly thinking about its partnership with Google ($GOOGL) on TPU chips. Google is scaling those chips and starting to sell them to outside customers. That helps Broadcom, which partners with companies to design their custom AI chips.
Broadcom wasn't as concentrated in AI semiconductors before. Now that business is becoming much more significant, and the latest quarter produced $13.67 billion of free cash flow.
I used to own Broadcom. I frankly sold too early because I didn't expect this kind of massive growth in its data center business.
Verdict: I think the stock is starting to look quite cheap. Trailing earnings are backward-looking, and so much of this growth has happened in the past few quarters.
3. TSMC: The chip shortage meets the factory bill
📊 Research TSM: Taiwan Semiconductor ($TSM)
Nvidia ($NVDA) keeps asking TSMC to build more. Nvidia knows it could sell more chips if it had them.
But TSMC is the company that has to spend the money on factories. It doesn't want to overbuild. That's one reason the whole AI buildout isn't even larger.
This is a more capital-intensive business, yet its net margin was 51.1% in the latest quarter. It manufactures chips for many companies, not only Nvidia.
Verdict: To me, an investment in TSMC is an investment in chip manufacturing in general. I think this sell-off makes it a pretty attractive opportunity.
4. Micron: Cheap earnings, or temporary earnings?
📊 Research MU: Micron ($MU)
Micron is charging more for products it already made. Memory is scarce, and data centers are willing to pay more for it. That's why the profits have exploded.
The low forward P/E reflects a real concern: the market thinks Micron is overearning, and those earnings could fall significantly when margins come down again.
The bull case is that AI has created lasting demand for memory. That looks true to me for next year and the year after. But the customers paying these prices are also trying to solve the shortage, including through designs and algorithms that use less memory.
There might be so much demand that those improvements happen and Micron is still fine.
Verdict: It looks like it could be a crazy value. The risk is what happens to profits if the shortage gets solved and Micron can't charge as much.
5. Comfort Systems: A tiny yield misses the growth
📊 Research FIX: Comfort Systems USA ($FIX)
Comfort Systems doesn't design or manufacture chips. Its HVAC business supports data center construction. It already had a large network of talent serving this market when demand exploded.
Revenue growth and wider margins together have done a lot for the bottom line.
The forward dividend yield was only 0.41% when I filmed. But the dividend's three-year compound annual growth rate was 58.74%. That's the part a low yield doesn't tell you.
My Take: I covered this stock during the April tariff sell-off and didn't buy it. I probably should have. After this decline, I think it looks cheap again based on earnings.
Earnings and Dividend News Links:
How did you like today’s newsletter?
📅 Keep Investing. Stay informed.
Zach
Founder, Dividend Data
P.S. Questions or suggestions? Reply to this email. I'd love your feedback!
Follow on YouTube | Listen on Spotify | Visit DividendData.com
Disclaimer: Dividend Data (Dividend Data LLC) is not a professional financial service. All materials released from Dividend Data (Dividend Data LLC) are for educational and entertainment purposes. Dividend Data (Dividend Data LLC) is not a replacement for a professional's opinion. Contributors to the Dividend Data (Dividend Data LLC) might have equities mentioned in the newsletter.