DuPont Analysis,Earnings Persistence,and Return on Equity: Evidence from Mandatory IFRS Adoption in Canada

This paper proposes four new models to forecast one‐year‐ahead return on equity (ROE) and change in ROE based on prior research in the DuPont analysis and earnings persistence, and also examines whether the persistence of ROE has improved upon mandatory IFRS adoption in Canada. Using the Granger causality test to establish the usefulness of additional explanatory variables in forecasting future earnings, I show that the DuPont components are useful in predicting one‐year‐ahead ROE, and that the persistence of ROE has decreased since Canadian firms adopted IFRS in 2011. This paper contributes to accounting research in two ways. First, it introduces a new approach to forecasting one‐year‐ahead ROE. Second, it sheds some light on the impact of IFRS adoption on reporting quality in Canada.