Modelling and Forecasting of Exchange Rate Pairs Using the Kalman Filter
Authors:
Open Access
- 15 November 2024
- journal article
- research article
- Published by Wiley in Journal of Forecasting
- Vol. 44 (2) , 606-622
Abstract
Developing and employing practically useful and easy to calibrate models for prediction of exchange rates remains a challenging task, especially for highly volatile emerging market currencies. In this paper, we propose a novel approach for joint prediction of correlated exchange rates for two different currencies with respect to the same base currency. For this purpose, we reformulate a generalized version of a bivariate ARMA model into a state space model and use the Kalman filter for estimation and forecasting of the underlying exchange rates as latent variables. With extensive numerical experiments spanning 18 different exchange rates (across both emerging markets, developing and developed economies), we demonstrate that our approach consistently outperforms univariate ARMA models as well as the random walk model in short term out‐of‐sample prediction for various exchange rate pairs. Our study fills a gap in the empirical finance literature in terms of robust, explainable, accurate, and easy to calibrate models for forecasting correlated exchange rates. The proposed methodology has applications in exchange rate risk management as well as pricing of financial derivatives based on two exchange rates.Keywords
Funding Information
- Engineering and Physical Sciences Research Council
This publication has 29 references indexed in Scilit:
- Forecasting exchange rates for Central and Eastern European currencies using country‐specific factorsJournal of Forecasting, 2020
- Predicting exchange rate returnsEmerging Markets Review, 2020
- Out-of-sample exchange rate predictability in emerging markets: Fundamentals versus technical analysisJournal of International Financial Markets, Institutions and Money, 2019
- Forecasting the exchange rate using nonlinear Taylor rule based modelsInternational Journal of Forecasting, 2019
- Exchange rate forecasting and the performance of currency portfoliosJournal of Forecasting, 2018
- Google Trends and the forecasting performance of exchange rate modelsJournal of Forecasting, 2017
- Forecasting exchange rates with a large Bayesian VARInternational Journal of Forecasting, 2009
- The performance of non‐linear exchange rate models: a forecasting comparisonJournal of Forecasting, 2002
- Prediction of future currency exchange rates from current currency futures prices: The case of GM and JYJournal of Futures Markets, 1994
- Empirical exchange rate models of the seventiesJournal of International Economics, 1983