期刊名称：International Journal of Computer Science and Network Security
印刷版ISSN：1738-7906
出版年度：2019
卷号：19
期号：3
页码：185-189
出版社：International Journal of Computer Science and Network Security
摘要：Markov Decision Processes enable researchers to analyze the dynamics of a stochastic process whose transition mechanism is controlled over time. In this study, the solution to a portfolio allocation problem using a Markov Decision Process (MDP) is formulated. The main subject of interest would be to find an optimal policy that minimizes the associated cost. The two challenges faced were, uncertainty about the price of assets which follow a probabilistic model and a large state/action space that creates it difficult to apply orthodox techniques to solve. Stocks data is obtained from Karachi Stock Exchange - 100 index (KSE-100) that have daily cumulative returns about 49.98% in 2012. It is found that all portfolios allocations achieve by Markov Decision Process, have better daily cumulative returns as compared to benchmark KSE-100 index. Therefore, it is concluded that Markov decision process is better approach to calculate assets’ allocation in designing stocks portfolios.
