AI Wars: 3 Dividend Stocks That Profit No Matter Who Wins

AI is the talk of the town these days. Google (GOOG) and Microsoft (MSFT) are locked into an all-out “AI war” that has seen the two companies compete over who can make the best AI chatbot. So far, Microsoft appears to be in the lead, having bought a $10 billion stake in OpenAI (a third of the company) and launched a chatbot demo that is wowing reviewers.

The competition between Microsoft and Google led to a steep selloff in Google shares. Google initially gained from the AI hype, rallying 7.3% on February 3, but it gave up the gains when a demo of its ChatBot produced an incorrect answer and Satya Nadella predicted that the company’s margins would drop.

It’s not clear right now who will “win” the AI arms race. Nadella’s claims about Google’s margins are certainly bold. Morgan Stanley (MS) recently estimated that working an LLM into Google would cause a $6 billion hit to EBIT, as LLMs tax servers more than search does. However, these estimates are, for now, just estimates.

How the “AI wars” will turn out is anybody’s guess. I’m pretty much neutral on it: I hold some Google, I used to hold Microsoft but sold it on valuation concerns. The price I sold MSFT at was much higher than today’s price, so my not holding it now isn’t a vote against buying it.

One thing I know for sure is that there are ways to get into the AI space without buying companies locked into fierce competitive struggle. If you’re willing to look at indirect AI plays–semiconductor companies, banks and others–you can find companies that profit off AI without being forced into a competitive steel cage match. In this article I will explore three dividend stocks that profit off of AI without facing the intense competition that characterizes the space.

Taiwan Semiconductor (TSM)

Taiwan Semiconductor is a dividend stock with a 1.91% yield. The company is probably the most obvious beneficiary of the AI wars that’s not named OpenAI. Microsoft, Google, and Meta (META) are all competing intensely with each other for eyeballs, advertiser dollars and other highly sought after prizes. This competition is the sort of thing that shrinks margins. It’s just the opposite with semiconductor foundries. Semiconductors are crucial to running the massive data centers that AI requires; the more advanced the AI, the more chips are needed to run it. TSM currently owns 60% of the foundry space, and if you look at how costly it is for a company to ramp up its foundry capabilities, you’ll see that TSM could easily hold its high share.

The source of TSM’s moat is the sheer cost of its operations. To make semiconductors you need ASML’s (ASML) EUV lithography machines, and they cost up to $300 million each. Just getting your hands on one of them is a major capital expenditure. So while companies may wish to compete with TSM, they will have to incur great costs to do so.

Take Intel (INTC) for example. It recently launched a foundry that competes with TSM, but it’s still nowhere near TSM’s market share. It has spent a lot of money on this business ($19 billion in 2021 alone), yet it’s still seeing its earnings decline. TSM by contrast was still growing its revenue last quarter and trades at just 14.5 times earnings and 8.61 times operating cash flow. A high moat AI-beneficiary stock that investors can trust.

JPMorgan Chase (JPM)

JPMorgan is a name you might be surprised to see on this list. Banks are not exactly known for being big players in AI, but this specific bank is. JPMorgan spends more money on AI research than any of its peers. It’s ranked #1 in AI adoption among 150 peer banks. And, its AI investments are paying off. Just recently, the bank announced that AI managed to do 360,000 hours’ worth of filing work in just a few seconds, saving JPM massively on labor costs. Obviously, this kind of thing will work wonders for JPM’s margins.

The story doesn’t end there. In addition to being a lender, JPM is also an investment bank, and the company’s AI investments could lead to benefits for that segment. When you underwrite deals and IPOs, it helps to understand the industry you’re working with, and JPMorgan would appear to have some “domain knowledge” that could be useful here. Will OpenAI go public? Will Tesla (TSLA) spin off its full-self driving unit to the public? Nobody knows the answers to those questions just now, but if such deals ever were to take place, it would help these companies to have a banker that knows about AI. So, JPMorgan could benefit from future deal-making in the AI space.

And what about JPMorgan’s overall business? As I’ve written in previous articles about banking stocks, banks can often benefit in periods when rates rise, like the one we’re in now. The effect isn’t guaranteed: inverted yield curves and recessions can offset the margin-boosting effect of high rates. However, JPMorgan is only trading at 11.7 times earnings and 4.55 times operating cash flow right now, so it looks like some risk is already priced in. Additionally, the stock has a 2.84% yield and a mere 33% payout ratio, so the yield is higher than that of the S&P 500 despite being very safe.

Micron Technology (MU)

Last but not least, we have Micron Technology. This is a dividend stock that I held in the past but sold recently. The yield on this stock is only 0.74%, and it has no dividend growth, making it far from a “high yield” dividend opportunity. But what it lacks in yield it makes up for in potential. Micron sells chips, like TSM does, but it only sells RAM and NAND Flash. RAM is short term memory; NAND Flash is a type of long term storage. The data centers that AI runs on require copious amounts of this stuff. This year, the prices of RAM and NAND Flash are going down, and that’s causing Micron’s revenue to decline. It guided for a loss in its upcoming quarter. These issues are among the reasons I recently sold my Micron shares. However, the stock remains relatively cheap, trading at 10.75 times earnings, 1.36 times book value, and 5.36 times operating cash flow, and it could see prices pick up steam again if AI leads to a surge in Data Center demand. I can’t vouch for this stock as strongly as I can the other two on this list, but it is a dividend stock that makes money selling to AI companies, so it merits inclusion.

The Bottom Line

The bottom line on AI is that it’s very much like a gold rush now, and history teaches that in gold rushes, you want to be selling the shovels, not prospecting. The competition for AI leadership is fierce, and nobody really knows who will win in the end. What we know with certainty is that chip companies will make money selling to AI companies, and that investment banks will profit off of deal making in the space. Perhaps these companies are safer investments than the companies that are directly throwing themselves into the competitive fray.