USING MARKOV CHAINS IN PREDICTION OF STOCK PRICE MOVEMENTS: A STUDY ON AUTOMOTIVE INDUSTRY

Abstract

Stock price prediction is on the agenda of most researchers based on the uncertainty of its nature. In past two decades the literature on the development of prediction models for stock prices has extended dramatically. These studies mostly focused on specific industries such as banking and finance petroleum manufacturing and automotive. In line with prior studies the aim of this study is also to provide a means for investors helping them predict price movements of stocks from automotive industry by using Markov Chains as it is one of the most commonly applied models. Automotive industry is not only a major and industrial force worldwide but also is a locomotive power that serves to many other industries. Daily closing stock price data of all 13 automotive companies listed in Borsa Istanbul (BIST) are collected for the calendar year of 2015. By defining three possible states (decrease increase and no change) individual state transition probability matrixes are formed for each company. Then using the probabilities provided with these matrixes different investment strategies are evaluated in the first five working days of 2016. According to the results of analysis it is concluded that applying Markov Chains generates a positive income or at least minimizes the loss.