University-Industry Engagement Week

U Toronto partners with Riskfuel to research AI as tool for derivatives trading and valuation


By David Schwartz
Published: October 27th, 2020

Riskfuel, a tech company focused on using artificial intelligence in the capital markets for derivatives valuation and trading, has entered a research partnership with the University of Toronto and its Department of Statistical Sciences focused on deep learning in financial modeling.

The partnership brings together some of the leading experts in AI-powered valuations for over-the-counter (OTC) derivatives, a $600 trillion market that includes interest rate swaps, credit default swaps, and structured products that aren’t traded on the stock exchange.

Riskfuel CEO Ryan Ferguson, one of the world’s foremost experts on applying machine learning to derivatives valuation, will lead his the partnership along with Professor Sebastian Jaimungal, director of the Masters of Financial Insurance program at the University of Toronto.

“We are excited to be partnering with some of University of Toronto’s world-class researchers on a huge, fast-developing new area of opportunity,” said Ferguson. “We have already demonstrated that AI in capital markets will be transformational, and working with Professor Jaimungal will allow us to stay at the forefront of innovation.”

The project is getting funding from the Natural Sciences and Engineering Research Council of Canada (NSERC). Riskfuel will share its industry expertise and ground-breaking applications, while the university team will explore new possibilities based on the latest research on machine learning and finance.

The project focuses on volatility surfaces, which use geometry to provide a window into the dynamics of current conditions in financial markets. Volatility surfaces are a necessary input for an AI model looking at derivatives, but these complex objects often confound traders and quantitative modelers.

“Riskfuel is a pioneer in developing [machine learning] tools in the financial markets, so they’re a natural partner,” said Jaimungal. “It’s exciting to be working with people who have an understanding of both industrial applications and AI. We will be working at the cutting edge, bridging academic explorations with practical, real-world applications. The potential here is wide and far-reaching.”

The research partnership will focus on real-time valuation for OTC derivatives, which are value to value because they can be contingent on interest rates, asset prices, or other economic indicators. Calculating all the possible outcomes is computationally intensive, and the current approach involves applying huge resources to provide overnight assessments of the previous day’s trading.

Riskfuel’s approach uses machine learning to speed up derivatives valuations and risk sensitivity calculations up to 1,000,000 times faster than current applications. Instead of overnight data, traders can have accurate information on derivatives values in real-time.

Source: GlobeNewswire

Posted under: University-Industry Engagement Week