- Stocks linked to artificial intelligence are soaring as investors assess the technology’s potential.
- Insider spoke to four tech fund managers about their ideas and approaches to the theme.
- While AI itself isn’t new, Wall Street is paying more attention in the wake of ChatGPT.
If artificial intelligence hasn’t quite taken over the world yet, it’s certainly captured Wall Street’s attention.
Last year, text-to-image models like Stable Diffusion and Da ll·E Mini went viral and got people thinking about what AI could do. This year, Open AI’s ChatGPT has attracted even more hype — and much more money — as Microsoft invested $10 billion in Open AI.
The reaction has included everything from weird art to predictions of doom, and it’s inspired a bona fide investor gold rush into the AI space. In the last 30 days, smaller stocks seen as AI plays such as SoundHound and Innodata have more than doubled in value. Big Bear has skyrocketed more than 400% so far this year.
That might feel familiar to investors who remember the hype around themes like the blockchain or the metaverse, which created hot buzzwords and hot stocks for short periods, but haven’t yet transformed the business world, to say the least. AI itself isn’t new technology, but the excitement around it has clearly reached a new level.
“Fads and themes are really common in technology,” mutual fund manager Matthew Moberg told Insider. He’s co-managed the Franklin DynaTech Fund for 18 years and handily outperformed most of his peers for the long term. “A lot of really great, impressive new technologies don’t have a cash register. They just don’t have a very good business model attached to them. Blockchain’s a great example of that.”
Still, tech investors feel that AI could be attached to an enormous cash register because it can help companies find valuable patterns in data that humans might never be able to see, or work much more quickly and cheaply in jobs like coding, customer service, and design.
Wedbush analyst Dan Ives, a famed tech-stock bull, says ChatGPT and AI could be worth about $150 billion in value to Microsoft alone.
In their “Big Ideas 2023” report, growth investors at Ark Investment Management say that if AI is adopted in 100% of possible uses in 2030, it could increase global labor productivity by $200 trillion. It says that amount is roughly equal to the projected GDP of the entire world that year.
That’s the bull case. But the bull case is never the only case. The debut of Google’s Bard was a fiasco, and Bing’s AI search has been riddled with errors and weird responses. Neither company is expecting big bucks from AI any time soon.
Insider spoke to Moberg and three other fund managers in tech about how they’re investing in AI’s potential at a time that a lot of money seems to be chasing a small number of stocks.
(1) ‘Picks and shovels’
One popular approach to investing in AI involves buying technology that numerous companies working in AI will have to use instead of trying to find a tiny “pure play” that might win big years from now. That’s the route recommended by Michael Grant, who manages Calamos Investments’ Phineus Long/Short Fund.
“Nvidia is the best investment on this theme today and it isn’t even close, because of the amount of compute power that you need for AI,” he said. “ChatGPT is currently being trained on about 25,000 GPUs, which is several hundred million dollars of Nvidia revenue.”
Grant’s fund has $935 million in assets, according to Morningstar, and over the three years ended February 21, its annual return of 11% more than doubled the index used as its benchmark, the Morningstar Global Target Market Exposure Index.
He adds that other chipmakers can benefit from that theme as well. Grant suggest that Advanced Micro Devices might be the biggest beneficiary after Nvidia, and Marvell Technology and Intel might also profit.
Grant says it’s too soon to pick an eventual champion in AI, but he likes the position Alphabet is in even though it’s had some notable AI-linked ups and downs recently. He says that Google’s parent company has been investing aggressively in artificial intelligence for years, and it has more access to data that can be fed to AI than anyone else.
“Google’s existing business is already at the foundation off where it can apply AI in a consumer-benefiting manner, which Microsoft is not,” he said. “This is Google’s future to lose.”
(2) ‘Choke points’
Ryan Jacob, the chairman and investment chief of Jacob Asset Management and a famed tech investor since the 1990s, says companies that want to use artificial intelligence in their businesses are going to have to work around a couple of potential problems that he calls “choke points.”
“AI will expose some issues that have been there before,” says Jacob. “These kinds of iterative AI applications use more computing power, there’s a higher degree of latency, they require massive amounts of storage.”
Jacob’s firm runs three stock funds and an ETF. The firm is known for whopping returns during bull markets and positive periods for tech, with much less favorable results in other periods. He says companies that help other firms address those problems will benefit for years. He says Cloudflare and MongoDB are two of his favorites.
“One of our larger holdings is Cloudflare,” he says, because it can help companies reduce wait times and save bandwidth. “A lot of the leading AI companies are using Cloudflare today, and we think that will increase.”
He says MongoDB stands out because it can help companies manage unstructured data, which is necessary in working with AI.
(3) Who benefits?
Michael Loukas is the CEO of TrueMark Investments, which launched its Tech, AI, and Deep Learning ETF three years ago. The $16 million ETF holds 21 stocks and has returned 14.4% in 2023. Loukas told Insider that he’s pursuing “hypergrowth,” and wants to buy companies that are taking a big lead in an important category.
“What business models, what companies, are going to be the greatest beneficiaries of these developments?” he asks. “Potentially the ones with the largest upside potential, or the longest runway, are the sophisticated users of artificial intelligence.”
His strategy is to hold a group of promising companies and shift his investments into winners as they emerge. His largest position is in Mobileye, a company that is working in autonomous driving technologies.
“Take a segment that is about to see an uptick in demand or trend or popularity or revenue and look for what company in that segment is going to be the greatest beneficiary and a potential category killer,” he said.
Loukas adds that most of AI’s potential is in business-to-business technology rather than in applications that consumers will use directly. For example, he says that companies like Schrodinger and AbCellera should win a lot of business with major drug developers.
(4) ‘Pools of money’
Moberg, of the Franklin DynaTech Fund, is a little more focused on finding major end markets. While AI may have many — he pointed to healthcare as an example — right now, they’re just potential.
“It’s really neat technology, but our job as investors is, ‘What are the applications? What is the business model around it? How do you invest and make money in it?’ That’s the part that remains to be seen,” he said.
He suggests that the smart way to invest in AI right now might just be waiting for those opportunities to reveal themselves, and investing in something that’s less trendy. That might be hard to hear based on the returns for tiny companies like Big Bear, but if the true potential of AI is years away from being realized, there will be busts and pitfalls between now and then.
“Even as someone who invests in innovation, right now you’re probably better looking in other areas than AI,” he said. “It’s just run up so much and kind of generically speaking so hyped and picked over.”
Grant, of Calamos, makes a similar point. While he named Nvidia as his favorite AI-related stock of the present, but notes that the hype has contributed to a 41% surge in its shares in 2023 even though AI-linked revenues are a very small portion of its business.
For companies like Alphabet, Microsoft, and Amazon, he says the present-day math is even less favorable, as they’re going to invest enormous sums of money in the technology over the next few years and won’t see notable returns for a while.
“The error that markets typically make is that technology shifts happen much quicker than the human shifts to ultimately support them,” he said. “It will take more than a decade for it to change in the way that will affect consumers and therefore many of the equity businesses.”