A wavelet network model for analysing exchange rate effects on interest ratesJournal of Economic Studies (2010)
AbstractAbstract Purpose – This research article discuss the effects of exchange rates on interest rates by using wavelet network methodology which is a combination of wavelets and neural networks. Design/methodology/approach – This paper employs wavelet networks to analyze the relationships between the financial time series. Empirically, the research examines the effects of foreign exchanges on the interest rates in Turkish financial markets by using daily USD/TRY rates and interest rates in Turkish Lira (TRY). Findings - The result indicate that wavelet network model is the most successful methodology among the alternatives such as Hodrick-Prescott filter, feed-forward neural network, wavelet causality, and wavelet correlation analysis in capturing the non-linear dynamics between the selected time series. Originality/value - The research results have both methodological and practical originality. On the theoretical side, the wavelet network is superior in modelling the causal linkages of the financial time series. For practical aims, on the other hand, the results show that the level of the effects of the exchange rates on the interest rates varies on the time-scale used. Wavelet networks shows that the causality relationship is strong in the short run, while the effect decreases in the mid-run.
- Computational intelligence,
- Wavelets networks,
- International finance,
- Exchange rates,
- Interest rates.
Publication DateSummer July, 2010
Citation InformationAlper Ozun and Atilla Cifter. "A wavelet network model for analysing exchange rate effects on interest rates" Journal of Economic Studies Vol. 37 Iss. 4 (2010)
Available at: http://works.bepress.com/alper_ozun/13/