Investing in artificial intelligence: The good, the bad and the ugly of AI
Crazy but true: If you want to take a high-risk gamble on a boom in “artificial intelligence” stocks, there’s an exchange-traded fund for that.
The Direxion Daily Robotics Artificial Intelligence & Automation Index Bull 2X Shares is a wild, leveraged fund that uses derivatives to try to give you twice the performance every day of the underlying AI and robotics sector.
That means twice the performance up — and down.
I am not making a recommendation. And this is not for the faint-hearted. (It may not be for anyone, but it’s certainly not for the faint-hearted.)
Try this: The fund has only been around for about five years. But during that time it has gone down 90%, up 700%, back down 80%, and then up 100%.
Urgh. Not on a heavy lunch.
After all that, it’s still down 50% from when it launched.
But don’t blame the fund. It’s doing exactly what it is supposed to do. “This leveraged ETF seeks a return that is 200% the return of its benchmark index for a single day,” the management company, Direxion, warns. “The fund should not be expected to provide two times the return of the benchmark’s cumulative return for periods greater than a day.”
It is not designed as a long-term hold.
As any kind of serious investment in this fund would be insane.
But as a trade — a bet on a quick AI mania?
Those willing to live dangerously might want to take a look. This week’s gigantic surge in Nvidia stock is just the latest signpost of a potential AI boom – or bubble. The good thing about stock market manias is that, for a brief period, they effectively rain free cash.
“Never let a bubble go to waste,” my late friend, London money manager Dan Bunting, used to say. And he’d add: “You make the most money from the worst stocks.”
The usual caveats apply: Never bet money you can’t afford to lose, never confuse a short-term trade with a long-term investment, and remember to get out when things go south.
Most of the tears from past bubbles came from those who broke those three rules. A short-term bet should never be confused with a long-term investment.
Bear in mind that the fund tracks the Indxx Global Robotics and Artificial Intelligence Thematic Index, which is not just about AI. The index has a very large exposure to companies involved in industrial automation, and in making robots, as well as AI itself (although AI chip giant Nvidia is the biggest single holding.) This differs from the Indxx Artificial Intelligence & Big Data Index, which is supposedly more focused on things like AI software and cloud computing.
Tejas Dessai, an analyst at fund company Global X, warns that he expects the two indexes to diverge in the years ahead.
But in the meantime, they’re marching pretty closely. Over the past year the two indexes have had a monthly correlation better than 90%.
Meanwhile, away from high-risk gambling… what about the rest of us?
If you’ve been watching the latest action and wanting to make long-term investments in the new AI boom, here are some things you should know before you do.
First, if you are confused about how to do so, you are not alone.
Even professional fund managers have no idea how to invest in AI – or even what it is.
Even mutual funds designed to invest in AI don’t really know how to do it.
“AI is a very challenging space to define,” Bryan Armour, analyst at Morningstar, tells me. He points out that the four specialist exchange-traded funds in the sector — Global X Artificial Intelligence & Technology ROBO Global Artificial Intelligence TrueShares Technology AI & Deep Learning and WisdomTree Artificial Intelligence and Innovation Fund — invest in totally different stocks.
He says there are only four stocks owned by all four funds – but 126 stocks that are only owned by one. “There’s certainly a wide range of opinions” on how to play AI, Armour says.
(The four on which the funds agree: Chip maker Nvidia, cloud computing company ServiceNow Amazon and communication technology company Twilio )
Other ETFs invest in AI plus robotics – like UBOT but without the leverage. These include Global X Robotics & Artificial Intelligence ROBO Global Robotics & Automation Index iShares Robotics and Artificial Intelligence Multisector and First Trust Nasdaq Artificial Intelligence & Robotics
Over the past 12 months most of these funds have a monthly correlation of 90% or better with the famous Invesco QQQ ETF which has been the benchmark for NASDAQ since before the dotcom bubble and tracks the Nasdaq 100 top stocks. In other words, if you owned the Nasdaq high-tech index fund QQQ you got more than about 90% of the action of the AI funds.
(LRNZ, an actively-managed fund with a concentrated portfolio of 20 to 30 stocks, had a lower correlation, about 75%.)
If you think you are going to be able to pick the long-term winners from AI, I have some used dot-com certificates to sell you.
During the internet bubble, the big winners were supposedly going to be companies like Yahoo, JDS Uniphase, Sun Microsystems, Nokia, and America Online.
And the top search engines back then included things like AltaVista, WebCrawler, and Ask Jeeves.
Apple was on hardly anyone’s list of top picks. Google was a small, little-known search engine. It wasn’t even public. Facebook didn’t even exist.
Meanwhile, for those betting on today’s surefire, blue-chip, can’t lose AI stock, the computer chip giant Nvidia, I have two words for you.
Cisco was the Nvidia of the dotcom bubble. Instead of chasing the online gold rush, it was going to make a fortune by selling the metaphorical “picks and shovels” to all those dot-com prospectors. In Cisco’s case this meant making the equipment on which the internet ran. In Nvidia’s case with AI, it means the computer chips.
At its peak, Cisco traded at 300 times the previous 12 months’ earnings and 20 times sales.
Nvidia today isn’t quite there, but it’s not a million miles away. The stock trades at 200 times trailing earnings, and also 20 times sales.
So what happened to Cisco?
Someone who bought the stock Cisco at the peak back then is still in the red 23 years later, even though during that time Cisco has quadrupled earnings. The success of the underlying business couldn’t offset the grotesque prices paid for the stock at the highs.
Meanwhile, someone who invested at the same time in a low cost global stock fund, like Vanguard Global Equity is up more than 400%.
Sure, Nvidia may fare totally differently. Who knows? But the higher the price you pay for a stock, the tougher the climb.
The good news about AI is that the actual benefits will filter through the whole economy. Companies across the board will be able to do more, and (maybe) better, with less. It should therefore be good for stocks (through higher earnings) and bonds (through deflation, or disinflation).
On the other hand, it’s probably not so good for us worker bees, who are expendable.
You want to bet on AI, maybe you need to look for unemployment plays. Or investments in trade schools. Someone will always need plumbers and electricians, right?