The Application of Wavelet Analysis in Financial Multiple Change Points Time Series

Abstract

In the dynamic financial market, the change of financial asset prices is always described as a certain random events which result in abrupt changes. The random time when the event occurs is called a change point. As the event happens, in order to mitigate property damage the government should increase the macro-control ability. As a result, we need to find a valid statistical model for change point problem to solve it effectively. Wavelet transformation method is introduced into financial market due to its convenience and simplicity. This paper proposes two methodologies which are Quandt-Andrews and wavelet transformation are to test the multiple change points stationary financial model in time series. We obtain the estimation of multiple change points, and compare the power of the two methods. From the real data analysis, the wavelet transformation method test is more efficient than Quandt-Andrews test.