Using the stochastic processes called Markov Chains, we sought out to predict the immediate future stock prices for a given company. We found the moving averages for the data and the grouped them into four different states of results. We then applied Markov Chain calculations to the data to create a 4x4 transitional probability matrix. Using this transition matrix we solved a system of equations and found four steady states that were variables that represented the probability that a stock price for a given day would fall into one of the four states. When we use this information we can apply our actual data to these equations and predict the next stock prices for the near future. We were able to successfully predict the next few days of stock prices using this method.
Worcester Polytechnic Institute
Major Qualifying Project
All authors have granted to WPI a nonexclusive royalty-free license to distribute copies of the work, subject to other agreements. Copyright is held by the author or authors, with all rights reserved, unless otherwise noted.