Evaluation of an Equation for the Canada-U.S. Real Exchange Rate

Abstract

Amano and van Norden first developed an equation for the Canada-U.S. real exchange rate in 1993, that empirically verified the existence of a relationship between commodity terms-of-trade variables, the interest rate differential and the real value of the Canadian dollar. Two years later, Lafrance and van Norden modified this equation, and reached a similar conclusion using quarterly data from 1972Q2 to 1993Q4. Using quarterly data retrieved from Federal Reserve Economics Data (FRED) and Statistics Canada, this paper revisits Lafrance and van Norden’s equation and empirically examines its feasibility from 1972Q1 to 2017Q1. First, I try to replicate the Lafrance and Van Norden’s (1995) results using the same sample period as they used (1972Q2 to 1993Q4). I also estimate the same equation with an extended sample period (1993Q4 to 2017Q1). The results show that the signs of the coefficients are consistent with those of Lafrance and van Norden, but their magnitudes are not; also, this equation has less explanatory power when the latest data are used. Next, after dividing the whole sample into five subsamples, each longer than the previous one, the major findings correspond to my expectation that the non-energy commodity terms of trade place upward pressure on the value of the Canadian dollar. However, the impact of the energy commodity terms of trade has shifted from depreciating to appreciating. Last, two modifications of the original equation aiming to improve the results demonstrate that the interest rate differential is a puzzling variable when it comes to explaining the movements of the real exchange rate; further research is needed to explain the unexpected finding that the interest rate differential no longer seems to influence the real exchange rate.