How Informative are Preliminary Announcements of the Money Stock in Canada?

We examine hypotheses about the relationship between provisional estimates and final values of M1, M2, and M3 and their growth rates in Canada, using monthly data and multiple revisions. Preliminary values cannot be viewed as final values plus an error (revision) uncorrelated with these but they are approximately unbiased forecasts of final values. The second difference in short-term interest rates is a leading indicator of revisions in M1 growth rates and revisions exhibit significant seasonality; hence preliminary values are not completely rational forecasts.