Working Paper: NBER ID: w4923
Authors: Richard C. Marston
Abstract: Two explanations are given for why nominal or real returns differ across currencies: foreign exchange risk premia and systematic (rational) forecast errors. This study reexamines three parity conditions in international finance, uncovered interest parity, purchasing power parity, and real interest parity, to determine the relative importance of these two factors. The study develops joint tests of the three parity conditions by relating nominal and real interest differentials and inflation differentials to the same set of variables currently known to investors. The study tests parameter restrictions based on knowing that risk premiums only affect nominal and real interest differentials, but not inflation differentials, while systematic errors in forecasting exchange rates only affect nominal interest differentials and inflation differentials, but not real interest differentials.
Keywords: foreign exchange risk; systematic forecast errors; uncovered interest parity; purchasing power parity; real interest parity
JEL Codes: F31; F36
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Risk premiums (G19) | Nominal interest differentials (E43) |
| Risk premiums (G19) | Real interest differentials (E43) |
| Risk premiums (G19) | Inflation differentials (E31) |
| Systematic forecast errors (C53) | Nominal interest differentials (E43) |
| Systematic forecast errors (C53) | Inflation differentials (E31) |
| Systematic forecast errors (C53) | Real interest differentials (E43) |