IFRS9 Expected Loss Models and SICR, an Alberta Perspective

Regulatory changes in accounting standards implemented in 2018 (rules IFRS 9) require that banks’ provisions (i.e., reserves set aside to cover future losses) are evaluated in a forward-looking manner, considering carefully-designed projections of the macroeconomic environment. Further, should a financial instrument experience a substantial risk increase, the methodology to calculate reserves must become more conservative. This project contributes to the development of compliant credit risk assessment methodologies in three ways. 1) evaluating the performance of the credit risk models currently used by the financial partner 2) evaluating state-of-the-art machine learning techniques to estimate expected credit losses and 3) evaluating the threshold at which the methodology used to calculate reserves must become more conservative.

Faculty Supervisor:

Valentina Galvani;Sebastian Fossati Pereira

Student:

Partner:

ATB Financial

Discipline:

Mathematics

Sector:

Finance and Insurance

University:

University of Alberta

Program: