Quantifying volatility clustering in financial time series

A quantitative method is introduced in this work to quantify and compare the volatility clustering behavior among various financial time series. In addition to financial markets, our approach can also be applied to other complex systems and we take the earthquake as an example to demonstrate the applicability of our approach. We further propose a toy model which can mimic the stylized facts in financial markets. This model could be interpreted as the accumulation effect of the news impact on the price fluctuation in a financial market and can be viewed as a first step towards understanding the complex market behavior.