1996•RePEc: Research Papers in EconomicsOpen access

Modelling the US$/A$ Exchange Rate Using Cointegration Techniques

Costas Karfakis, ANTHONY J. PHIPPS

Open full text 0 citations

Abstract

Recent evidence indicates that Australia's real effective exchange rate, its terms of trade \nand a long-term real interest rate differential form a cointegrating relationship. This paper \nuses this evidence to analyse the nominal US$/A$ exchange rate. The US$/A$ rate is found \nto be cointegrated with the terms of trade and relative price levels. However, interest rate \ndifferentials appear to add nothing to this long-run relationship. Estimated error correction \nmodels suggest that there is a substantial two-way relationship between nominal exchange \nrate changes and changes in the terms of trade. This evidence indicates that the small, open-economy \nassumption of exogenously given terms of trade may be inappropriate when modelling \nmovements in the US$/A$ exchange rate. Changes in a long-run interest rate differential, \npossibly reflecting differences in expected inflation rates, contribute significantly to \nan explanation of short-run changes in the nominal exchange rate.

Open-access reader

About this research paper

What this paper is about

Recent evidence indicates that Australia's real effective exchange rate, its terms of trade \nand a long-term real interest rate differential form a cointegrating relationship. This paper \nuses this evidence to analyse the nominal US$/A$ exchange rate. The US$/A$ rate is found \nto be cointegrated with the terms of trade and relative price levels. However, interest rate \ndifferentials appear to add nothing to this long-run relationship. Estimated error correction \nmodels suggest that there is a substantial two-way relationship between nominal exchange \nrate changes and changes in the terms of trade. This evidence indicates that the small, open-economy \nassumption of exogenously given terms of trade may be inappropriate when modelling \nmovements in the US$/A$ exchange rate. Changes in a long-run interest rate differential, \npossibly reflecting differences in expected inflation rates, contribute significantly to \nan explanation of short-run changes in the nominal exchange rate.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Recent evidence indicates that Australia's real effective exchange rate, its terms of trade \nand a long-term real interest rate differential form a cointegrating relationship. This paper \nuses this evidence to analyse the nominal US$/A$ exchange rate. The US$/A$ rate is found \nto be cointegrated with the terms of trade and relative price levels. However, interest rate \ndifferentials appear to add nothing to this long-run relationship. Estimated error correction \nmodels suggest that there is a substantial two-way relationship between nominal exchange \nrate changes and changes in the terms of trade. This evidence indicates that the small, open-economy \nassumption of exogenously given terms of trade may be inappropriate when modelling \nmovements in the US$/A$ exchange rate. Changes in a long-run interest rate differential, \npossibly reflecting differences in expected inflation rates, contribute significantly to \nan explanation of short-run changes in the nominal exchange rate.

Key concepts: Cointegration, Economics, Exchange rate, International Fisher effect, Fisher hypothesis, Inflation (cosmology), Econometrics, Differential (mechanical device)

Related papers

Back to paper searchBrowse research topicsOriginal source
Modelling the US$/A$ Exchange Rate Using Cointegration Techniques — Research Paper | ScholarLens