ChatGPT, the hugely popular Artificial Intelligence (AI) chatbot, might be able to write impressive articles in seconds. But it’s so far unable to accurately predict stock prices, much to traders’ dismay.
For instance, when asked to forecast the current banking crisis’ winners, it failed to provide specific names. However, it listed five segments that could benefit, cautioning: “it’s important to note that a banking crisis can have significant negative impacts on both banks and customers and predicting ‘winners’ is not necessarily appropriate given the potential negative externalities. However, in response to your question, here are the potential beneficiaries of a banking crisis:
- Large, well capitalized banks…
- Government regulators and policy makers…
- Insurance companies…
- Debt collectors…
- Digital and mobile banking…”
The chatbot went on to explain why these sectors could gain from the crisis that rattled global markets in recent days, noting that lenders with strong balance sheets will likely emerge stronger from the shake-out while regulators could win “increased authority, public support and funding.” Insurance firms and lending firms, meanwhile, will benefit from greater demand for their products while digital banking can draw new customers.
ChatGPT, which uses current and past Web data to yield results, also failed to specifically answer the following query: “What will be First Republic Bank’s (FRB) stock price in a year?” replying “I am sorry but as an AI language model, I don’t have the ability to predict stock prices or financial outcomes,” noting that “there are many factors that can affect the stock prices of companies.”
When trialing ChatGPT-4, the bot’s upcoming version, in a response to a question about UBS’s share price in five years, I received a more cogent answer: While noting “it’s difficult to predict what the price of a stock, such as UBS, will be in five years, given that the stock is currently trading at $14 per share, it is safe to say that the price will be higher at some point in the future.”
UBS, the Swiss banking giant that just absorbed ailing rival Credit Suisse to help end the financial crisis, was up 12% to 19.42 CHF on Tuesday afternoon trading, a day after the bot was queried.
Show me
Still, many market watchers say the bot and other AI systems are not yet equipped to reliably predict the stock market prices and that they are not yet using them to hammer out strategies.
“You have to show me it works before I believe it,” said Sam Stovall, chief investment strategist at brokerage CFRA. “So far, it’s not accurate so to think it can predict stock prices is wishful thinking.”
Jane Edmondson, who heads ETF index provider EQM, said the technology is promising, however, and that chat bots such as GhatGPT or Google’s (GOOG) Bard, will likely refine their prediction abilities to eventually perfect them within 10 years. Her team is considering pairing one of her quant models with an AI engine to enhance future investment outcomes.
“It’s a very interesting area to build future products and indexes,” she noted.
Live trading
Oxford University researchers are taking it further, however, recently launching an AI program that can forecast real-time stock prices. The model can project how share prices move in real-time trading and has had an 80% success rate for the equivalent of about 30 seconds of live price movements.
“We effectively have a model trained to make forecasts at smaller horizons,” Stefan Zohren, associate professor at the Oxford-Man Institute of Quantitative Finance and co-author of the research, recently told Bloomberg. “We can feed this information back into itself and roll forward the prediction to arrive at longer-horizon forecasts.”
The institute reportedly exploited principles of “natural-language processing” to analyze liquidity data across buy-and-sell limit order books to build the model. Meanwhile, those interested in longer time frames can visit CoinPriceForecast.com which uses an AI model to list future prices for a slew of digital currencies such as bitcoin and stocks like Tesla (TSLA).
New ETF
Meanwhile, ETF providers are rushing to exploit the frenzy behind AI Chatbots, a technology loosely known as ‘generative AI,’ by launching a new vehicle dubbed Roundhill Generative AI & Technology ETF (CHAT). Pending regulatory approval, it will bankroll firms at the ‘forefront’ of the space, said Chief Strategy Officer David Mazza.
He conceded it will take some time for AI to accurately predict future share prices. Eventually, however, the technology will add “firepower” to existing fundamental and technical analysis tools at investment firms looking to raise their game, also by mining historical data. When that happens, trading as we know it could change radically.
“Trading will no longer be an entertainment business,” Stovall said. “It will become pretty mundane and a lot of the financial TV shows will probably go away.”
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
Image and article originally from www.nasdaq.com. Read the original article here.