A Duquesne Capital veteran will use artificial intelligence to help investors anticipate cryptocurrency fluctuations through his fintech company TOGGLE AI.
TOGGLE AI, co-founded by Jan Szilagyi and Giuseppe Sette, uses artificial intelligence to track price shifts in stocks, commodities, and fixed income. The company announced Tuesday that it would start analyzing over 400 cryptocurrencies in addition to major currencies like Bitcoin and Ethereum, which were previously included in TOGGLE’s analytics.
Szilagyi, the firm’s chief executive officer, told Institutional investor that the appeal of artificial intelligence resides in its ability to analyze, process, and present millions of data points at an inhuman speed. This means that platforms like TOGGLE can present users with predictions for over 35,000 securities, an impossible feat for a single human analyst.
“AI is able to process a large amount of data and then process certain patterns that might be useful,” he said.
Szilagyi and his co-founder Sette, president of TOGGLE AI, spent the bulk of their careers in asset management. Szilagyi started his career as a quant trader at Stan Druckenmiller’s Duquesne Capital (Druckenmiller would later become TOGGLE AI’s first investor). After receiving his doctorate in quantitative finance, Szilagyi spent the next ten years in fundamental global macro investing — studying the data and understanding the drivers for individual asset performance.
Before launching TOGGLE AI, Szilagyi and Sette worked as co-chief investment officers of global macro strategies at Lombard Odier Investment Managers. There, it became obvious to them that the amount of available data vastly exceeded their ability to analyze it.
“The magnitude and the speed at which we were able to get both macro and micro data became overwhelming,” Szilagyi said.
So they launched TOGGLE AI in 2020. The platform, which Szilagyi calls a “smart list,” delivers analytics incorporating factors like earnings expectations, sales for individual companies, momentum for price in a particular share, geopolitical events, and fiscal policy developments — anything that may affect the future of the securities in a client’s portfolio.
If any of these factors poses a substantial risk or opportunity to the portfolio, TOGGLE will flag it. For example, if analysts’ expectations become incrementally less positive for a certain stock, the system will highlight this and alert investors.
“TOGGLE will have given you and delivered an entire bit of analysis that will say, ‘Look, we’ve noticed this kind of deterioration, say, 27 other times for this stock. This is how the stock has typically behaved in the aftermath over the past weeks and months, and we think it’s relevant that you take a look at it,’” Szilagyi said.
He said the platform is particularly helpful for investors with large portfolios who may neglect certain positions, because the AI is programmed to take all positions into account at once, leaving nothing up to chance.
Once TOGGLE gained momentum and a framework was established for identifying portfolio risks and opportunities, cryptocurrencies became the obvious next step, Szilagyi said. Crypto was just beginning to enter the mainstream investing world when the company was launched in 2020, but demand has since surged, and the amount of related data has exploded — allowing TOGGLE to generate meaningful analysis based on historical data.
“The amount of data that’s available about the blockchain is mind-boggling,” Szilagyi said.
In addition, because of cryptocurrencies’ relative newness, Szilagyi said there’s no widely-established or long-held opinions about how to trade them. It’s a space primed for innovation.
“This is where computers really shine because they’re able to go at such a faster pace compared to an individual trying to do this on their own,” he said.
As cryptocurrencies become more mainstream, Szilagyi said that is greater demand for crypto analysis.
“If you had developed a system like this five years ago, there would have only been a handful of truly dedicated crypto followers that would have been interested in this,” Szilagyi said. “As the ecosystem matures, I think the demand for tools like this one becomes more essential.”
