Finding Hidden Pattern of Financial Time Series Based on Score Matrix in Sequence Alignment

Abstract

This paper applies sequence alignment method of bioinformatics to financial analysis to find hidden pattern from financial markets. Results of simulation suggest that sequence alignment method can be used to identify key points to inset, delete and replace data in time series, to find lead-lag relationship between two time series, and to analyze matching patterns. We further propose a new score matrix named similarity-oriented matrix which is designed based on the characteristics of financial time series, and apply it to China’s stock market. The empirical analysis verifies the validity of our proposed score matrix, and tests the sensitivity for different threshold values of symbols definition.