First published: https://www.utoronto.ca/news/foreseeing-future-oil-prices
How do you predict the future trading price for a barrel of oil? You may be more likely to succeed if you have a statistician, a financial expert and computer scientist working together.
Waterfront International Ltd. (WIL), a quantitative finance research firm specializing in developing computer-based statistical trading strategies, challenged University of Toronto students to develop an automated trading strategy for oil prices – in only 48 hours.
“The programming – or hacking – community is known for its intensive, problem-solving approach,” said Helen Kontozopoulos, coordinator and instructor of entrepreneurship‐related initiatives in the department of computer science. “Quantathon was designed to challenge U of T students from a diverse range of computational studies.”
“We use technology to solve complicated finance problems and rely on people with backgrounds in computer science, statistics, mathematics and finance,” said Bob Suriwka, Waterfront International’s chief executive officer. “We approached the departments of computer science, statistical sciences and Rotman to create with us a contest that mirrors what we do in our business.”
The multidisciplinary teams, 27 in all, were given a quantity of physical crude oil. They then applied futures contract prices for oil and current market rates to maximize their return-to-risk ratio and forecast the oil price. The groups used heuristics, simulation, linear regression, optimization, machine learning and time series analysis to design their oil trading strategy.
“The teams worked really hard at this hackathon kind of question, but as it applies to a different field, in finance,” said Suriwka.
Quantathon’s question proved highly challenging. At the end of the second day, only 18 of the 27 teams were left standing with completed solutions. A shortlisted group of teams were given an opportunity to present their results to a panel of judges.
Department of statistical sciences doctoral students Luhui (Luke) Gan, Tianyi Jia and Zhenhua Lin, took home the top $7,500 prize courtesy of Waterfront International.
“We study a lot of theory in finance, statistics and computer science,” said Gan. “We’re very interested in real-world applications and wanted to see how theory can be used.”
“Nowadays there’s a big data area, where people want to able to find the answer within and you need a computer program to help you find the answer,” said Jia, who credits teammate Lin for bringing his prior programming experience to the team.
“The problem was very cool and realistic,” said Lin.
“It was fun. It was hard. It was open-ended,” said runner-up, Mufan (Bill) Li, a fourth-year undergraduate student in the Edward S. Rogers Sr. Department of Electrical & Computer Engineering (ECE), who formed a team with ECE classmate Mengye Ren and fourth-year physics student Japinder (JP) Nijjer.
Li and Nijjer said they were attracted to Quantathon based on their work experience in quantitative finance. Their second-place finish netted the team $5,000.
Third place of $2,500 went to a team of students from UTSC: Lucas Huang (finance), Veronica Ng (statistics) and David Szeto (computer science).
“We were impressed by the amount of turnout we had for the competition, the quality of student work and the amount of time put in,” said Suriwka.
“Not everyone can win, but everyone did a great job.”
Waterfront International plans to expand Quantathon in 2016 to additional Ontario cities.
Contact Helen Kontozopoulos, for more information firstname.lastname@example.org
I'm the Co-Founder at ODAIA.ai & an Adjunct Professor, at the Department of Computer Science, University of Toronto, Canada