Ong, Phang Cheng (2012) Behavior of share prices by using first order Markov Chain Model. Universiti Malaysia Sabah. (Unpublished)

Text
ae0000002866.pdf Download (822kB)  Preview 
Abstract
In this modern era, fluctuation economic growths make people worry with their future. Therefore, investment acts as an essential role in life for overtaking such crucial problems and aid in development in a nation. Invest in stocks is one of ways for investment. Hence, technique of forecasting has been proposed for investors as a guideline for analyse the behavior of shares and to maximum their profits. Markov chain is a stochastic process and a kind of forecasting method which will be implemented in this study. Five years historical data from the Finance Yahoo and The Star Business for AirAsia Berhad had been collected to estimate missing value by using polynomial order 9 and to forecast. In order to forecast the share prices, matrix of order 3x3,5x5,7x7 and 9x9 is built. Next, transition probabilities for different dimension of matrices are constructed by using the method of equal interval, equal probability and standard deviation. Probability forecasting is done by applying the state probability vector from ten criteria and transition probability matrix. Lastly, (Mean Absolute Percentage Error) MAPE is used to test the forecast ability in order to obtain the best model for forecasting. As results of this study, method of standard deviation is the best method compare with the method of equal interval and equal probability. Besides that, matrix 3x3 from method of stand deviation has obtain the lowest MAPE which is 8.60% when applying with the first condition of initial probability vector which is using duration of five years data.
Item Type:  Academic Exercise 

Keyword:  Markov chain, forecasting method, Mean Absolute Percentage Error, standard deviation 
Subjects:  Q Science > QA Mathematics 
Department:  SCHOOL > School of Science and Technology 
Depositing User:  ADMIN ADMIN 
Date Deposited:  02 Jun 2016 08:37 
Last Modified:  30 Oct 2017 11:28 
URI:  https://eprints.ums.edu.my/id/eprint/13308 
Actions (login required)
View Item 