Which AI is better for picking stocks?

AI chatbots can write a poem, do your homework, draft demands, and maybe even take your job, if the hype is to be believed. Can they handle your investments too?

While the use of artificial intelligence in the realm of financial advice is nothing new (“auto advisors” have been around for years, some of which use AI), chatbot technology is quickly becoming more accessible to investors. individual.

Google’s Bard and Microsoft’s chatbot, powered by ChatGPT and integrated into its Bing search engine, can interact with users in simple language and can engage in interactions that seem surprisingly human.

To test the investing capabilities of Microsoft’s and Google’s respective products, we challenged each to pick two stocks, a growth stock and a value stock, and see how they fared over a three-week span compared to each other and with a human. We used Investopedia’s stock market simulator, so there was no real money changing hands.

Neither product was specifically designed to offer financial advice, and both framed their choices with caveats, just like we did. Investopedia is not trying to recommend AI as an investment tool, or promoting the stocks the robots picked.

We start by asking each bot the same question: If you were to buy one value stock and one growth stock to hold, which would provide the higher return?

The selections:

Bard began by offering the definitions of growth stocks and value stocks, and suggested Walmart as a value option, calling it a “well-established company with a strong track record of profitability.” For a growth stock, he singled out its creator’s competitor Microsoft, praising the tech giant, in uncannily similar language, as “a well-run company with a strong track record of innovation.”

For value, Bing suggested Verizon and Walgreens, refusing, even after urging, to narrow it down to just one. He provided data to accompany his selection. Unfortunately, the data was from 2021 (possibly because the underlying ChatGPT technology only used data up to 2021) and was not accurate: it said Verizon reported 5% year-over-year revenue growth in Q3 2021, when it actually reported a growth of 4.3%. For a growth stock, he recommended Shopify.

This human reporter picked stocks at random, choosing Lockheed-Martin because the military, at the time, had just shot down a Chinese spy balloon and two unidentified objects in US airspace, apparently a golden opportunity for aerospace defense companies sell their products. . For a value stock, I chose Campbell Soup because of the conventional wisdom that the affordable food maker would do well in the economic downturn many economists were predicting.

The results:

Although all portfolios rose over the course of the test, the clear winner was Bard, whose combined picks rose 5.15% over three weeks, beating out both Bing and its human opponent, as well as outperforming the S&P 500.

The experiment doesn’t really challenge conventional wisdom about stock selection. Experts generally discourage retail investors, human or robot, from buying individual stocks and instead advise a diversified portfolio. The market is difficult to predict, and it is not even ruled out that asset prices move completely randomly.

However, what was obvious was that while the bots could answer the questions in a natural way, they clearly did not have much information about the financial markets, at least not yet.