Front Matter Page
Research Department
Authorized for distribution by Michael Mussa
Contents
Summary
I. Introduction
II. Previous Literature
A. Static productivity-based models
B. Static productivity-based models with rigidities
C. Dynamic models
III. Theoretical Model
IV. Time Series Approaches
A. Time series econometric methodologies
B. Data
C. Time series results
V. Panel Regression Approaches
A. Panel regression methodology
B. Estimating the cointegrating relationships
C. Panel regression results
D. Panel cointegration results
VI. Trend Exchange Rates: Productivity-Based Models vs. PPP
VII. Conclusions
Tables
1. Previous Estimates of Productivity Coefficients
2. Single-Equation Estimates of Four Variable Model
3. SUR Estimation Results: Determinants of Real Exchange Rate
4. ECM and PL Estimates of Deviation from Trend
Figures
1. FM-NLS and SUR Estimates of the Rate of Reversion
2. NLS and SUR Estimates of the Tradables Productivity Coefficient
3. FM-NLS and SUR Estimates of the Nontradables Productivity Coefficient
4. FM-NLS and SUR Estimates of the Foreign Government Spending Coefficient
5. United Kingdom
6. France
7. West Germany
8. Italy
9. Canada
10. Japan
11. Belgium
12. Denmark
13. Netherlands
14. Norway
15. Sweden
16. Finland
17. Australia
References
SUMMARY
This paper investigates the determinants of the real exchange rate using a panel of disaggregated data for the OECD countries. It also marries two literatures—one that uses panel data to measure relationships between changes in exchange rates to changes in the determinants, and another that uses cointegration techniques to measure the long-run relationship between the level of the exchange rate and the level of the determining factors. The previous panel studies cannot account for deviations from long-run trend levels, while the extant literature using time-series cointegration techniques can only intermittently detect and measure posited relationships. Estimating the relationships in levels is an interesting enterprise because it allows one, in principle, to calculate trend real exchange rates.
After surveying the previous literature, a dynamic fixed-factors Balassa-Samuelson model of the real exchange rate is used to motivate the empirical exercise. In examining this problem, we exploit recent developments in the econometric analysis of nonstationary variables in panel data. The results indicate that under certain assumptions it is easier to detect cointegration in panel data than in the available time series; moreover, the rate of reversion to trend is estimated with greater precision. The half-life of a deviation from long-run equilibrium is on the order of four to five years.