Image source: Getty Images
Generative AI bots have already taken the world by storm, providing users with answers in lightning-quick time. But are they any good for selecting great stocks to own for passive income? I thought it might be fun to find out.
What I discovered was actually quite worrying.
The usual suspects
My weapon of choice for this little experiment was ChatGPT and my question was simple: “What are the five best stocks to buy for passive income?”
In only a few seconds, I was provided with a list of names: M&G, Aviva, Legal & General (LSE: LGEN), Phoenix Group and HSBC.
It was immediately clear that these companies shared a few attributes.
All of the above are members of the FTSE 100 — the index tracking the UK’s biggest businesses by market cap. All are likely familiar to most Foolish investors as well. At least a few are household names.
Perhaps most interestingly, all offer dividend yields of over 6% at the current time. Contrast this with the 3.5% offered by the index as a whole and it starts to look like AI could be a very useful tool when it comes to stock selection.
Spot the difference?
By now, however, you’ve probably noticed something: all of ChatGPT’s picks specialise in providing financial services of some sort. We’re talking the biggest insurance companies and wealth managers around, not to mention a £150bn bank.
I find this frightening.
Imagine if a fresh-faced investor put every penny they had to work in these five stocks. Such overconcentration is very risky, especially if sentiment around the UK (or global) economy sours.
Now, I’m certainly not suggesting that ChatGPT recommended a bunch of absolute stinkers. But it does show the danger of relying on anything other than old-fashioned research to pick investments. After all, the bot had no knowledge of my financial goals or time horizon. Nor did it provide any reasons for why it thinks these stocks are ‘the best’.
Safer picks
Pushed to select one stock to buy from the list above, however, I’d probably opt for Legal & General. Unlike some of ChatGPT’s picks, it has a pretty solid record of consistently distributing dividends over time, at least since the dark days of the Great Financial Crisis. As things stand, analysts are forecasting another bump to the total payout in FY25.
A growing payout is something I particularly look for since it implies that a business is in decent health. This means more to me than going for only the highest-yielding stocks — something the bot didn’t seem fussed about.
However, one needs to be wary. Legal & General’s dividend looks set to only just be covered by expected profit. That can’t carry on for too long or a cut will likely follow.
For this reason, it would be prudent to consider getting exposure to other, unrelated parts of the market to balance out the risk here. Defensive sectors like consumer goods, utilities and healthcare are particularly appealing.
What I’ve learned
Looking ahead, I don’t doubt that I’ll employ AI in some form for researching investments. But I would never dream of using it to build a portfolio on its own.
No one cares more about my money than I do. And that includes the seemingly all-powerful, all-knowing (but really rather reckless) ChatGPT.
