Jan Szilagyi, CEO and co-founder of software company Reflexivity, shared his interesting insights into the use of artificial intelligence for stock trading. During an appearance on CNBC's SquawkBox show earlier today, Szilagyi shared that top hedge funds are using AI software as an add-on to their existing tools, adding that AI has made several correct stock market predictions this year.
The executive, whose company develops analytical engines that overlay large language model (LLM) interfaces, believes that given enough computing power and training, artificial intelligence can beat human intelligence in stock trading.
Szilagyi said at the beginning of the interview that Reflexivity’s ultimate goal is to “create an autonomous investment analyst.” As part of this process, the company is currently providing technology that enables investors to integrate data from disparate sources onto a single platform to streamline decision-making.
By asking questions to the "engine," users can "rely on the engine to find the data and then analyze it," the CEO said, reducing the time required for such operations from two hours to two minutes.
He outlined that Reflexivity's current AI trading system is designed to "find parallels from the past to give you an idea of what price patterns might look like in the future." These "similarities" can be "based on the past 12 or 15 events, selecting securities that will correspond to what you are looking for," involving the current environment or a set of economic parameters or other parameters that may induce certain stocks to deliver better returns than others.
Szilagyi's clients "are primarily top hedge funds that use it as a smart overlay on top of all the different data sources they have access to," he shared.
He also explained in detail how artificial intelligence correctly predicted several trends in the stock market in 2024, but missed others. Szilagyi outlined, "If you look at the market as a whole, for example, it was very correct in calling the peak in July, and it was also very correct in calling the false rebound after that. After the peak, the AI" again correctly predicted the bottom. "
However, the executive added that his software turned bullish "a little prematurely" ahead of the Fed's latest meeting, which guided for two rate cuts in 2025 instead of the previous four.
However, Szilyagi doesn't think the model went wrong here because its output is "always a probability assessment." According to him, these assessments are essentially models "saying, OK, know there's a 70 percent chance that this is going to happen. There's a 30 percent chance that it's not going to happen."
When asked when artificial intelligence could completely replace humans in stock trading, Szilyagi controversially said, "It's hard to say there's anything so special about human intelligence that we should be better at trading." He believes, "If you can give the system so much computing power, and the intelligence continues to improve, then at some point, it is not a crazy idea to think that the system will be better than humans at trading."
He concluded from this that "we may need five to ten years" to replace humans. This does not mean that AI will completely replace humans, as some industries, such as private markets, will be harder to target due to a relative lack of training data. "
Compiled from /ScitechDaily