Study shows ChatGPT and AI may have a future as your portfolio manager
Lionel Bonaventure | AFP | Getty Images
The proliferation of artificial intelligence programs like ChatGPT and Alphabet’s BardAI has already made big waves in financial markets, and a new study suggests that one day those programs may be able to trade in those markets on their own.
The investment industry has long used algorithms and quantitative trading programs in an effort to boost profits, but recent breakthroughs in AI have opened up new possibilities. While a possible future where AI moves money on its own is likely several years away, a preliminary paper from two academics in South Korea suggests ChatGPT is already a better portfolio manager than throwing darts.
The paper – “Can ChatGPT improve investment decisions? From a portfolio management perspective” – found that ChatGPT’s ability to choose from a set of assets outperforms random selection on measures of risk-adjusted return and diversification.
“Retail investors, especially those who may be illiterate or misinformed, could benefit from the democratization of portfolio management,” wrote the paper’s authors Hyungjin Ko and Jaewook Lee. “Additionally, professional portfolio managers can improve their productivity by focusing on more important tasks while receiving assistance from ChatGPT in selecting diverse assets for a given portfolio.”
The researchers said that the initial results suggest that ChatGPT can be used as a “co-pilot” for investors, but not a “prophet”. It’s not too far off from the robo-advisors already in use for many retail brokerages. It also reflects the plans of at least one financial firm, as Morgan Stanley is already testing OpenAI-powered chatbots with its financial advisors to help with investment decisions.
The experiment, which has not yet been peer-reviewed, went like this: The researchers created a universe of 20 large-cap stocks from different sectors of the US market, and five from each of the categories of cryptocurrencies, commodities, currencies and bonds. Created assets. ,
Then, the researchers ran 10,000 simulations and asked ChatGPT to select different numbers of available assets to build a portfolio.
A backtest of the data showed that ChatGPT’s selections created a more diversified portfolio with less correlated assets than random selections.
And then, during the period January 1, 2022 to January 31, 2023, the ChatGPT portfolio outperformed even the random selection portfolio on a risk-adjusted basis.
For example, in a portfolio of four assets, the ChatGPT portfolio had an average expected return of -13.3% versus -23.3% for random selections and a better Sharpe ratio, which is a measure of risk-adjusted return.
Stocks and bonds declined sharply in 2022, so negative return results are not surprising. For comparison, the S&P 500 fell more than 14% during this time period.
stock chart iconstock chart icon
The S&P 500 fell during the time involved in this experiment.
This study does not mean ChatGPT is ready for prime time and will soon be running hedge funds. Citi analyst Chris Montagu said in a note to clients that the results were “reassuring” but cited the small asset universe, short time frame and “easy-to-beat baseline model” as reasons to be cautious about widespread adoption. cited.
But as language-learning AI models continue to improve and investors increasingly migrate to low-fee products, AI portfolio managers could one day become a reality.
— CNBC’s Michael Bloom contributed to this report.