Publications and Research

In this study the authors compare the information content of alternative monetary aggregates with respect to total spending in the economy, using data for Canada. The analysis considers 46 monetary measures, about half of which constitute conventional summation aggregates, while the remainder are superlative indices of monetary services based on the Fisher Ideal formula. The methodology follows a natural sequence in which the information set used to predict total spending is gradually expanded, beginning with lagged total spending itself, followed by the monetary aggregates, then other financial variables such as interest rates, exchange rates, and stock prices. Finally, the information loss due to monetary aggregation is tested by decomposing a broad and informative aggregate into its components, within a multivariate indicator model. All of the competing models are tested against one another, using the encompassing principle of Hendry and Richard (1982), and then evaluated by means of standard stability tests, one-period-ahead in-sample forecasting, and post-sample simulation.