Recommender Systems for Investing

(1) The Desjardins Quantitative Strategies department manages a set of internally developed systematic investment strategies. There are two main products; global equity strategies (developed and emerging countries), as well as alternative strategies using futures on global stock indices, resources, interest rates and currencies. The team owns a proprietary technology platform that has been developed over a period of more than 10 years and is actively used for fund management.
(2) The universe of available financial securities and the amount of financial, economic, social, governance, and environmental data is a big data problem. Identifying relevant factors and data to make optimal investment recommendations under constraints is
computationally intensive. Further, the recommender system should be interpretable or explainable so that it is understood what risk factor contribute to the inclusion or exclusion of a particular investment.
(3) The main benefits, due to its impact on our portfolios and to accelerate the adoption of machine learning approaches internally at Desjardins, would be the construction of a novel stock recommendation system based on fundamental and technical properties of investments, and interactions with the macroeconomic environment or environmental considerations.

Faculty Supervisor:

Cody Hyndman;Frédéric Godin

Student:

Partner:

Desjardins Gestion Internationale d'actifs (DGIA)

Discipline:

Mathematics

Sector:

Finance and Insurance; Artificial Intelligence

University:

Concordia University

Program: