Using ChatGPT to predict stock prices
Corona Del Mar, CA
Howdy Friend,
Wouldn't it be nice to just type into ChatGPT "what price will Apple stock be tomorrow" and out spits the answer? Yes, that's a fantasy, but I just read a research report with a novel use for ChatGPT's sophisticated language model.
(Title: Can ChatGPT Forecast Stock Price Movements? Return Predictability and Large Language Models)
They used ChatGPT's API to analyze news headlines and rate them as good, bad, or neutral. For example, take a look at this headline: "Rimini Street Fined $630,000 in Case Against Oracle" |
ChatGPT rated the headline as "good" saying:
"The fine against Rimini Street could potentially boost investor confidence in Oracle’s ability to protect its intellectual property and increase demand for its products and services."
Pretty smart answer eh? Now let's skip to the results:
Their backtest was only 15 months, but showed that the average stock declined by 0.46% the next day if ChatGPT determined bad news had come out. Good news didn't seem to be very profitable, but that could be a side effect of that 15-month backtest occurring during a bear market. Much of the profits on the short side could also be due to the bear market.
Overall, I find the strategy novel and interesting, but it appears to be on par with Twitter sentiment strategies I've seen...which means it severely under-performs strategies that use TAP data.
True Asset Price (T.A.P) data shows actual mispricings of ETFs (to the tune of billions per day). |
Sentiment is just someone's opinion.
I'll take objective truth over subjective opinion any day of the week. Speaking of objective and measurable, something remarkable happened here at Portfolio Boss...my team just told me that they were able to speed up the 3,500 computer cores that make up The Boss SuperAi - by 300%-500%!
That means our on demand cloud computing fees are a fraction of what they used to be. I could buy a fleet of Ferraris for what The Boss was costing every month.
But instead, I'm going to pass those savings to YOU! That's right, we're having a good 'ol fashioned salebration on what could be your best money-makin' opportunity ever.

|
Disclaimer: The results listed herein are based on hypothetical trades. Plainly speaking, these trades were not actually executed. Hypothetical or simulated performance results have certain inherent limitations. Unlike an actual performance record, simulated results do not represent actual trading. Also, since the trades have not actually been executed, the results may have under (or over) compensated for the impact, if any, of certain market factors such as lack of liquidity. You may have done better or worse than the results portrayed.
Responses