Quelques résultats empiriques relatifs à l'évolution du taux de change Canada/États-Unis
This paper explores the extent to which factors other than commodity and energy prices may have contributed to the Canadian dollar's depreciation since the early 1970s. The variables considered include among others budgetary conditions and productivity. The approach involves a long-term determination, using cointegration methodology, of variables that may have played a major role in the behaviour of the real exchange rate. The authors conclude that, while growing indebtedness in Canada has contributed to the Canadian currency's depreciation against the U.S. dollar, it explains only 20 per cent of that decline during the 1990s.
Non technical summary
Amano and van Norden (1995) have developed a model for determining the real Can$/US$ exchange rate that has proven itself particularly robust over time, in terms of statistical significance and stability. This model consists of both short- and long-term components. The long-term component takes the form of a term of error correction arising from a linear relationship between the real exchange rate, commodity prices (excluding energy), and the price of energy. The short-term dynamic is essentially induced by the spread between interest rates in Canada and the United States. In this paper, the authors attempt to examine the extent to which factors other than commodity and energy prices may have contributed to the Canadian dollar's depreciation since the early 1970s. The variables considered are selected on the basis of theoretical or earlier empirical criteria, and include the productivity gap between Canada and the United States, the gap in the ratio of public spending to GDP, the gap in income per capita, net foreign assets, and the gap in the ratios of public debt to GDP.
Of these variables, only the gap between debt-to-GDP ratios in Canada and the United States would appear to provide any additional information. Including this variable improves the global specification of the model and its predictive power, especially for the period relating to the early 1990s. Although commodity price variable continue to dominate real exchange rate dynamics over the entire sample, debt assumes ever greater importance from the mid-1980s, when the Canada-U.S. indebtedness gap began to widen. The portion attributable to indebtedness tends to fade towards the end of the 1980s, but reasserts itself in the mid-1990s, when the indebtedness gap (as a proportion of GDP) reached record levels. The indebtedness gap at that time could explain as much as 20 per cent of the depreciation of Canada's real exchange rate, but this percentage starts to decline towards the end of 1997, as the country's public finances strengthened. It is noteworthy that the price of energy becomes steadily less significant as the estimation period is lengthened. Thus, the commodity price index and the Canada-U.S. indebtedness gap alone provide an almost complete explanation of real exchange rate behaviour over the period 1974-1998. This observation is consistent with various empirical tests that point to a continuous drop in the predictive power of energy prices in several macroeconomic relationships.