A signal processing model for time series analysis: The effects of international F/X markets on domestic currencies using wavelet networks

Abstract

This paper proposes a powerful methodology wavelet networks to investigate the effects of international F/X markets on emerging markets currencies. We used EUR/USD parity as input indicator (international F/X markets) and three emerging markets currencies as output indicator (emerging markets currency). We test if the effects of international F/X markets change across different timescale. Using wavelet networks, it is found that the effects of international F/X markets increase with higher timescale. This evidence shows that the causality of international F/X markets on emerging markets should be tested based on 32-64 days effect. We also find that the effects of EUR/USD parity on Turkish Lira is higher on 9-16 days and 33-64 days scales and this evidence shows that Turkish lira is less stable compare to other emerging markets currencies as international F/X markets effects Turkish lira on shorten time scale. Besides it is found that Russian ruble is mostly affected from international F/X market in the long time according to wavelet networks analysis. Copyright © 2008 Praise Worthy Prize S.r.l. - All rights reserved.