Nowhere for Investors to Hide From AI Hype
4 min read
02 Jun 2023, 05:59 PM IST
Even if you can look past the froth, there are few places to invest
PremiumInvestors thinking of putting money into artificial intelligence need to apply a little real intelligence first. The gold rush begun by chatbot ChatGPT is quickly turning into a mini-bubble—and there’s no need to look far back in history to see how quickly bets on hot new themes can put your portfolio into the deep freeze.
This week, chip maker Nvidia was briefly worth more than $1 trillion, thanks to investor excitement about demand for its chips created by so-called “generative AI," the technology behind ChatGPT. Smaller AI-related stocks have also soared as investors search for the big winner from the latest fashionable theme.
The shift is easy to understand. ChatGPT was launched on an unsuspecting public in November and hit 100 million users in two months, making it the fastest-growing consumer application ever. Aside from the bad jokes, bad poetry and clunky LinkedIn-style articles it can churn out, it is being used and abused by students, lawyers, CEOs and programmers.
Naturally, investors want to find the companies that will win from what is being talked up as the biggest technological development since the internet. There are two problems. The first is that AI might once again be overhyped. The second is a dearth of places to invest, leading to some extraordinary stock surges.
AI has certainly been overhyped before. I spent time as an undergraduate studying neural networks and expert systems during a wave of AI hype in the early 1990s, before one of the many AI winters made the effort pointless. The hype reached new highs in 2010, when IBM’s Watson system was reported to be capable of beating humans at quiz show “Jeopardy!" IBM’s stock surged as Watson went on to win the contest in 2011, and was launched commercially—but IBM shares have plunged more than a third since peaking shortly thereafter.
The hype cycle might finally have reached the point where AI will be obvious to consumers, rather than providingbehind-the-scenes improvements, such as to Alphabet’s Google Translate or camera facial focus. It’s certainly caught the public imagination in a way that previous AI breakthroughs didn’t—as well as catching the eye of regulators.
Unfortunately for the would-be AI investor, there are few listed companies dedicated to AI. Who will be the winners from AI developments? The obvious candidate is Nvidia, whose chips are the picks and shovels of the AI gold rush. But Nvidia stock is already up 160% this year, and it trades at 44 times estimates of 12-month forward earnings—even after analysts doubled their estimates. Nvidia has to not only win from the AI rush but win really, really big—merely to justify where the shares are already.
AI investors have piled into the rest of Big Tech for their pick-and-shovel equivalents in the form of cloud services, or because they are also investing heavily in AI. As a result, the biggest tech stocks have broken away from the rest of the market, where more S&P 500 stocks are down than up this year.
The struggle to find places to invest shows up in the AI exchange-traded funds. The biggest is BOTZ, the Global X Robotics and AI ETF, which had the 19th-largest inflows of any ETF tracked by FactSet in the past month, despite not even making the top 100 by size. It has the same problem as everyone else, eased somewhat by having the majority of its money in robotics, rather than AI. The biggest holding, naturally, is Nvidia, but this year’s best performer is a tiny company called C3.ai that’s up 210%.
C3.ai is a classic for investors who like themes. It is on its fourth iteration, having tried business models in carbon emissions, energy management and the “internet of things"—each with a company name change—before alighting on AI. Investors focused on growth will love it, as revenues have almost tripled in four years (although its shares plummeted after slightly disappointing forecasts on Wednesday). Investors who care about the business should look at what it’s spending to get those sales: It recorded losses of 36 cents for every dollar of sales in the year to April 2019, before its IPO, but in the past year lost just over $1 for every dollar of sales. At worst they are trying out the old joke in real life, losing money on every sale but trying to make it up in volume. At best, C3.ai’s ability to scale subscriptions into profit is unproven.
The odds are that the AI hype is just that, once again. As with the much larger dot-com bubble, or the more recent mini-bubbles in 3-D printing, clean technology, cryptocurrencies, cannabis stocks and SPACs, it will last as long as the hype lasts, then dissipate. Someone, somewhere, will surely make big money out of AI, and many firms may improve productivity. But for sure not all the currently-tipped stocks will be winners, and perhaps not any of them.
A final word: If you’re thinking of buying a fund that uses AI to pick the stocks for you, don’t get your hopes up. The oldest of these is AIEQ, the AI Powered Equity ETF, which has used IBM’s Watson since 2017. Since its launch, it has returned 34% including dividends, against 82% for the S&P 500. Oh, and it is underweight Nvidia.
"Exciting news! Mint is now on WhatsApp Channels 🚀 Subscribe today by clicking the link and stay updated with the latest financial insights!" Click here!
