Discovering Pattern Associations in Hang Seng Index Constituent Stocks

Abstract

The problem of finding patterns in financial time series has been tackled by systematic observations of trends, statistical analysis or the use of artificial intelligence techniques in trend analysis. These techniques are more for the discovering of patterns in data rather than the understanding of association relationships between the discovered patterns. As time series patterns often overlap with each other, identifying and discovering association relationships among them can be very challenging. To tackle these problems, we propose here a method to determine if there exists any association relationship between two sequential patterns in a financial time series. The method is based on the use of machine learning techniques and has been tested with data from Hang Seng Index (HSI) constituent stocks. The results reveal that there is statistical evidence of association relationships between some of the stocks whereas there is no evidence for such a relationship between some others. We conclude that the price behavior of these HSI stocks is easier to understand than that of the HSI index.