Using ChatGPT to Pick Stocks
Topics: AI and Picking Stocks
The Dangers of Using AI
Everyone is talking about AI these days. When used correctly, it is an amazing tool that can be used for endless possibilities. Although, I recently came across an article named ChatGPT can pick stocks better than your fund manager. From those few words, I already knew that I was getting to get angry over this article.
Spoiler Alert: I was.
Anna Cooban, the writer of the article, outlines that ChatGPT selected a portfolio of stocks and “gained 4.9 percent while 10 leading investment funds clocked an average loss of 0.8 percent” between March 6 and April 28. My first concern over this experiment is that it took place over a span of 53 days.
I am more curious how this experiment will go over a longer period of time, such as over a five-year time frame. This level of AI is new, so I understand why studies over the long term have not been performed. Although, believing that AI will consistently outperform leading investment funds based on this experiment is dangerous. Especially based on just 53 days, the AI could have “lucked” out with some of its picks.
Additionally, prompting AI to “pick stocks” is a risky approach as well. If the AI just spits out a list of five stocks without much reason, then there is reason to be concerned. There is never a one size fits all approach when it comes to investing, and that must be known when leveraging AI. For example, AI could pick a portfolio of riskier securities while you may want a portfolio of safer securities.
Furthermore, I am curious if ChatGPT picked higher-risk assets which led to earning a higher return. A common idea in investing is that taking on higher risk equates to a higher return. Or alternatively, taking on higher risk equates to a higher loss as well. The reason you could earn a higher return is that you are accepting more risk as an investor and will be compensated for doing so.
The combination of small sample size and not knowing why ChatGPT picked certain securities are the major red flags from this experiment. Although, I do not want to completely denounce AI for stock research. Given that it is accessible and free, it can be useful research for finding stocks or looking at financials. In other words, it can be a starting point for research and should never be used as the “end all be all” answer on if you should select a stock.
As I have heard many times during my academics, picking stocks can be “an art and a science”. As with everything that involves AI, the “art” aspect is entirely removed. Be conscious of that if you ever use AI for stock picks.
My Experience - Chat GPT / Notion AI
Chat GPT: Welp, looks like I will never (potentially) beat the market.
Notion AI: Then I moved to Notion AI and asked “recommend me five stocks”. Very ambitious for Notion AI to select 5 stocks that are all included in the top 10 of market cap for American companies.
Sources:
-CTV News: https://www.ctvnews.ca/business/chatgpt-can-pick-stocks-better-than-your-fund-manager-1.6386348
-Reddit: https://www.reddit.com/r/technology/comments/13ao8i6/chatgpt_can_pick_stocks_better_than_your_fund/
