USING CONTINUOUS WAVELET TRANSFORM TO ANALIZE MULTIVARIATE FINANCIAL TIME SERIES
USING CONTINUOUS WAVELET TRANSFORM TO ANALIZE MULTIVARIATE FINANCIAL TIME SERIES

Summary/Abstract: The excessive investors' interest in one market sector may cause bubbles that are indications preceding financial crises. The recent financial crisis started as a housing bubble, therefore it was not an exception. An increasing volatility of markets is a common characteristic of all crises. Such processes evolve simultaneously and thus generate co-movements off inancial time series. Understanding if their co-movement is causal or not and how contagion spreads is of great interest. There are several methods to model co-movements, from basic ones (the correlation coefficient), to more advanced ones (Vector Autoregressive models, Cointegration analysis, GARCH models). An alternative and very effective approach is by means of Wavelet analysis. Wavelets combine the analysis of financial data in the frequency domain and in the time domain. This paper focuses on the Continuous Wavelet Transform and attempts to apply the wavelet correlation and the wavelet coherence to study the co-movements of financial time series.

• Details
• Contents