Alphabet (google): We asked Google’s AI for 10 stocks to invest in, here is the result

(BFM Bourse) – After having already done so with ChatGPT at the start of the year, on July 13 BFM Bourse asked Bard, Google’s conversational robot, ten stocks in which to invest over a three-month horizon. How have these values ​​performed since then?

Important disclaimer: we have engaged in a pure theoretical exercise with ChatGPT and Bard. In no way do we recommend investing on the basis of advice given by artificial intelligence, nor choosing such a short horizon of three months. The usual advice (become well informed about the securities in question, diversify your portfolio, focus on the long term) prevails.

Will Bard do “better” than ChatGPT? As we had already done with OpenAI’s conversational robot at the beginning of the year, we asked Google’s to give us stocks to invest in over a three-month horizon.

with five titles on Wall Street and five titles on the Paris Stock Exchange.

A difference with ChatGPT can already be observed: Bard responds directly to this request and gives 10 actions without problem, although with a (useful) caveat.

“Disclaimer: I am not a financial advisor and this is not financial advice,” he warns. “Please note that this is only a list of 10 stocks that you may consider investing in. It is important to do your own research before making any investment decisions,” recommends he elsewhere.

In comparison, we had to insist with ChatGPT in January to achieve our goals. And on several occasions, including this week, we tried to repeat the experience but without success. “I’m sorry, but I cannot provide specific investment advice or stock recommendations because stock prices can be very volatile and subject to various factors that can change quickly,” the chatbot replied to us from OpenAI.

>> Access our exclusive graphic analyses, and gain insight into the Trading Portfolio

Very consensual values

Concerning the values ​​retained, Bard did not worry, like ChatGPT for that matter. On Wall Street, its choices focused on the same values ​​as the OpenAI robot. Namely Apple, Microsoft, Amazon and Nvidia. Only difference, on the last title: Bard cited Tesla against Bank of America for ChatGPT. But the two robots spontaneously mentioned well-established tech groups that feature among the highest market capitalizations in the world.

For the five Parisian stocks, Bard cited well-known CAC 40 groups, namely Danone, Axa, TotalEnergies, LVMH and Airbus. We nevertheless notice a certain sectoral diversification. ChatGPT had also given TotalEnergies and Airbus, but had also set its sights on Sanofi, Orange and BNP Paribas, thus avoiding the luxury sector.

In the end, how did the portfolio Bard cited perform? Did it perform better than ChatGPT over the same period (we kept the 10 January values ​​cited by ChatGPT)? To determine the variation of these 10 values, we assumed an equal investment for each security in the “portfolio” constructed from the actions evoked by each robot.

Answer: the stocks cited by Bard could not avoid the slump that has affected the stock market in recent months. This is much less true for ChatGPT.

Good performance in Paris

On Wall Street, the five stocks cited by Bard fell by an average of 2.8%, with tech groups having been weighed down by the rise in bond yields which is weighing on the valuation of growth groups. This is a little better than ChatGPT’s five (-3.5%) simply because Tesla fell less than Bank of America over the period (-5.4% against -9%).

The result is a little more interesting in Paris. Thanks to Totalenergies (+18.7%), which was supported by the rise in oil prices and a successful investor day, as well as Axa (+7%), Bard’s five Parisian stocks did not show an average decline over the period by only 0.58%. The average is, however, weighed down by LVMH (-22%), while luxury is experiencing a stock market downturn due to a disappointing recovery in China and a normalization of demand. LVMH also disappointed with the publication of its third quarter activity this week.

For its part, ChatGPT achieved a clearly positive performance in Paris, of 6.95%. Like Bard, Totalenergies is pushing the figure upwards, and, moreover, the absence of LVMH is good. In addition, the three values ​​cited only by ChatGPT performed well. Orange gained 4.6%, thanks recently to a positive rating from Bank of America, BNP Paribas gained 5.4%, driven by good half-year results, and Sanofi gained 11.5%, while Berenberg switched to buying the stock and that risk aversion in the market may have benefited stocks considered defensive, such as pharmaceutical groups.

Luck?

Ultimately, the gross variation (excluding exchange rate effects and commissions, management and/or financial intermediary fees) of the ten Bard stocks stands at -1.69% since July 13. This is a slightly smaller decline than that of the CAC 40 over the same period (-2.8%) and the S&P 500 (-3%). ChatGPT, thanks to its Parisian values, did better, with an increase of 3.4%.

Let’s assume that you invested 10,000 euros in the 10 securities (1,000 euros on each security) of Bard and ChatGPT on July 13, you would currently have 10,749 euros with the first and 11,083 euros with the second, excluding financial intermediary fees and commissions. Why are these capital gains higher than the gross performance? Quite simply thanks to exchange rate gains on American stocks, the dollar having appreciated by more than 5% since July 13, the date on which the euro almost reached its 2023 peak against the euro. Since then, risk aversion has benefited the greenback and battered the eurozone currency.

Let’s be realistic: whether it’s Bard or ChatGPT, their better performance than the market is certainly due to luck and we can only advise you not to try the experiment we did.

In a post published on LinkedIn in February, Stephane Renevier, analyst for the financial portal Finimize, explained that ChatGPT could potentially help dissect the strengths and weaknesses of a stock, but always by verifying and deepening its analysis.

“ChatGPT is far from perfect: it was trained on a pre-2022 dataset. Its content is often misleading or just plain wrong. And it doesn’t understand the content it produces at all,” he wrote . [pour faire simple de la surperformance sur un portefeuille, NDLR]”ChatGPT will not allow you to discover hidden gems. It will not allow you to directly generate alpha

. You won’t become a superstar analyst by using it,” he insisted.

Always check data from an AI

According to a questionnaire conducted by the Certified Financial Planner Board of Standards (CFPB), a grouping of two American organizations, in August, 31% of individual investors (out of a total of more than 1,150 adults) declared (unfortunately) to be “at risk”. comfortable” with the idea of ​​taking financial advice via generative AI tools, without verifying that advice and information with another source. Which you obviously shouldn’t do.

“Social media and generative AI show promise as information sharing tools, but they can never replace the expertise and client knowledge of a financial planner,” warned Kevin Keller of the CFPB. “Verifying data with a reputable third-party source (…) is the best way for consumers to ensure they stay on track to achieve their financial goals,” he added. Advice that is very valid on the stock market…

Markets.com’s Zachariah Walker pointed out in May that ChatGPT cannot analyze new data or information as soon as it appears, meaning “ChatGPT has the potential to produce unreliable results.” “Which may be fine for a little fun, but becomes a serious risk when asked to deal with important financial matters that have serious implications for the user who is relying on the algorithm to function productively,” he warned.