Stocks selected by ChatGPT far outperform UK fund managers

Between March 6 and April 28, a dummy portfolio of 38 stocks gained 4.9 per cent while 10 leading investment funds clocked an average loss of 0.8 per cent, according to an experiment conducted by financial comparison site finder.com, CNN reported.

It wouldn’t “be long until large numbers of consumers try to use (ChatGPT) for financial gain,” Jon Ostler, Finder’s CEO, said in a statement earlier this week.

Over the same eight-week period, the S&P 500 index, which tracks the 500 most valuable companies in the US, rose three per cent. Europe’s equivalent, the Stoxx Europe 600 index, ticked up 0.5 per cent in that time.

A typical investment fund pulls together money from multiple investors, and is overseen by a fund manager who decides how to invest that money, CNN reported.

Finder’s analysts took the 10 most popular UK funds on trading platform Interactive Investor as a benchmark for assessing the performance of the ChatGPT-generated fund. Funds managed by HSBC and Fidelity were among those selected.

Discover the stories of your interest

The analysts asked ChatGPT to select stocks based on some commonly used criteria, including picking companies with a low level of debt and a track record of growth. Microsoft, Netflix and Walmart were among the companies selected.While major funds have used AI for years to support their investment decisions, ChatGPT has put the technology in the hands of the general public, with the potential to guide the decisions of retail investors, CNN reported.

A survey of 2,000 UK adults conducted by Finder last week showed that eight per cent had already used ChatGPT for financial advice, while 19 per cent said they would consider doing so.

Yet a much bigger 35 per cent said they would not consider using the chatbot to help them make decisions about their money, CNN reported.

In a study published in April, researchers at the University of Florida found that ChatGPT could predict the stock price movements of specific companies more accurately than some more basic analysis models.