Stock price fluctuations and information

Abstract

This thesis analyzes stock market based on concepts of information theory such as entropy, joint entropy, conditional entropy and mutual information. The analysis was done with real stock market data of Samsung Electronics, Hyundai Motor Company, Hyundai Mobis, SKT, POSCO, Korea Electric Power Corporation, total six items which cover the seven-year period from January 1998 January 2005. The mutual information measures the amount of information that one random variable contains about another random variable. It indicates how closely two stocks are related. By using the correlation coefficient as a measure of randomness we obtain a similar result as the above. Although the absolute values of the two measures are different, the patterns are similar. It is found that the correlation coefficient and the mutual information react in different aspects according to the random variables.