The Probe Algorithm with Information Map: A New Dynamic Method for Correlation Analysis

A time lag effect cannot be ignored while information spreading in one market or between some related markets. This effect makes the market price of the associated time series have similar fluctuations. The former methods like VAR model, Copula Function and so on can propose the entity relationship between two time series for a certain period, however, the relationship may be changing all the time. In order to solve this problem in a dynamic view, a “Probe Algorithm” with Information Map is proposed in this paper. Empirical analysis using stock market data is also given, and the results show that, in both macro and micro levels, Probe Algorithm can effectively improve the reliability of the correlation analysis. Moreover, using the Information Map, the evolution paths of the relationship can be figured out clearly