Real exchange rates and real interest rates once again: a multivariate panel cointegration analysis

Abstract

This paper employs multivariate panel cointegration techniques to re-examine the empirical relationship between bilateral real exchange rates and real interest rates. The results from a panel of 1470 quarterly observations on Canada, France, Germany, Italy, Japan, UK, and USA over the period 1977 to 1994 indicate the absence of any long-run relationship between the two variables.

Notes

¹ See Taylor (1995) for an insightful discussion on this issue.

² See Pedroni (1999).

³ See Isard (1982).

⁴ It may be noted that following the work of Hooper and Morton (1982), researchers have included the terms ccbal and ccbal* or ccbal/gnp in order to capture the familiar idea that the expected long-run real exchange rate is a function of some ‘fundamental’ factors.

⁵ See Im et al. (2003).

⁶ The IPS test statistic is an average of the Augmented Dickey–Fuller (ADF) statistics over cross-sections. See Dickey et al. (1986).

⁷ See Phillips and Moon (1999).

⁸ See Pedroni (1998).

⁹ Till date the applicability of similar techniques has mostly been limited to the case of a single regressor. See Pedroni (1997).

¹⁰ Shiller and Perron (1985) demonstrated that the total length of the sample period rather than the frequency of observations is the important factor when examining the long-run properties of time-series.

¹¹ Two alternative measures of inflation (4-quarter change in the CPI and 12-quarter centre moving average of the CPI) were used as well. The results were similar.

¹² The augmented Dickey–Fuller (ADF) test when applied individually to each country indicated the same: most of the variables are not level stationary but all variables are I(1) in each country.

¹³ See Pedroni (1999).