Taylor Rules in the Quarterly Projection Model

In recent years, there has been a lot of interest in Taylor-type rules. Evidence in the literature suggests that Taylor-type rules are optimal in a number of models and are fairly robust across different models. The reaction function in the Bank of Canada's Quarterly Projection Model (QPM) is an inflation-forecast-based (IFB) rule. A number of studies have suggested, however, that the optimality of IFB rules is very model-specific. Given this and concerns about model uncertainty, it seems logical to assess the performance of Taylor-type reaction functions in QPM. Therefore, we compare QPM's IFB rule with a simple Taylor rule as well as with two rules that include open-economy elements. Overall, our results suggest that Taylor-type rules do not perform well in QPM compared with the base-case IFB rule, since they are associated with significantly higher variabilities of inflation, output, and interest rates. However, of the Taylor-type rules considered, we find that a simple rule with a coefficient of 2 on the contemporaneous inflation gap (versus 0.5 in Taylor's original rule) and a coefficient of 0.5 on the output gap is the most appropriate. Furthermore, the gains from using open-economy rules seem to be limited.