Content area
Full Text
Can monetary and financial variables predict factor markets are about to become too slack or too tight for non-inflationary growth? Predictions of recessions and booms are useful because they can help firms plan inventories and expansion of capital and labour, and make counter-cyclical policy more effective.
The definition of recession is fairly standard. Lacking a definition of boom we test three alternative definitions based on deviations from potential gross domestic product. We find financial variables are more useful for predicting recessions than booms. A negative yield spread between long-term Government of Canada bonds and the 90-day commercial paper rate is the best predictor of recessions up to five quarters ahead, but a positive spread only helps predict booms one quarter ahead. Growth in real M1 is marginally useful in anticipating recessions. No variable predicts booms beyond two quarters.
Methodology
To determine whether alpha^sub 1^ in (2) is statistically different from 0, we can use a statistic analogous to the t-statistic in linear models for the in-sample results. The parameter is statistically significant at the 95 per cent level if the t-statistic exceeds 1.985.
We tested 20 monetary and financial variables, including nominal Canadian and U.S. interest rates, interest rate yield spreads, real stock indexes, monetary aggregates and stock performance indicators. Selection criteria were historical availability, measurement accuracy and timeliness. For example, interest rates and stock prices are good candidates since they are available over long horizons, are not subject to revision and are available daily. For our recession study the data set spans the quarterly period from the first quarter of 1957 to the fourth quarter of 1996. Availability of potential GDP constrained the range of the boom model from the fourth quarter of 1968 to the fourth quarter of 1997. Sources are provided in the Appendix.
The recession periods taken from Cross (1988) and Macklem, Paquet and Phaneuf (1995) are presented in Table 1. One can debate the precise beginning and ending of recessions, but the selected dates capture all periods of major economic stagnation over the past 40 years.
The definition of boom periods is more ambiguous. Our working definition of a boom is an economy at risk of overheating because it is operating at or above its full-employment GDP potential. This...