Premium
Forecasting exchange rates for Central and Eastern European currencies using country‐specific factors
Author(s) -
Jaworski Krystian
Publication year - 2021
Publication title -
journal of forecasting
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.543
H-Index - 59
eISSN - 1099-131X
pISSN - 0277-6693
DOI - 10.1002/for.2749
Subject(s) - exchange rate , robustness (evolution) , econometrics , economics , random walk , principal component analysis , inflation (cosmology) , exploit , monetary economics , financial economics , computer science , statistics , mathematics , computer security , theoretical physics , gene , biochemistry , chemistry , physics
This study builds on two strands of the literature regarding exchange rates—developing methods to forecast them and attempting to find a link between exchange rates and macroeconomic fundamentals (i.e., addressing so called “exchange rate disconnect puzzle”). We propose looking separately at its global component (common for all the currencies) and the local component (country‐specific one) instead of modeling and forecasting the exchange rate directly. We demonstrate that in the last few years, local factors have been gaining importance in shaping the exchange rate returns for the Polish Zloty, Hungarian Forint, Czech Koruna, and Romanian Leu. We further show that the main drivers of the local component of exchange rate returns are the future values of the gross domestic product growth rate and consumer price index inflation. Using principal component analysis combined with linear regression, we exploit this tendency for forecasting purposes. Our novel approach yields superior results compared to the random walk in out‐of‐sample forecasting exercise at horizons of 1 month to over a year in the case of Central and Eastern European currencies. The results withstand the sensitivity and robustness analysis.