Quah, Sim Peng (2008) Forecasting Maybank's share prices using Markov Chain and Arma Model. Universiti Malaysia Sabah. (Unpublished)
|
Text
ae0000000913.pdf Download (1MB) | Preview |
Abstract
Data that is used to forecast in this study is closing prices of Malayan Banking Berhad (MAYBANK). The trend and changes of share prices are studied and observed. The closing prices from 3 January 2000 to 29 December 2006 are used for data analysis. The closing prices from 3 January 2007 to 31 January 2007 are used for accuracy checking. Bursa Malaysia does not operate on weekend and public holidays. Thus, missing values exist and are estimated using Cubic spline interpolation. Five-state Markov chain, sevenstate Markov chain, nine-state Markov chain and eleven-state Markov chain are being analysed. The forecast performances of different states of Markov chain are being compared and found that seven-state Markov chain obtained the best results. The better forecasting tool in forecasting share prices is investigated. ARMA model is better forecasting tool in forecasting share prices compared to Markov chain. The significance difference between model with and without missing values is being investigated. It is concluded that there is significance difference between models with and without missing values. Mean absolute percentage error (MAPE) is used in comparisons among the models.
| Item Type: | Academic Exercise |
|---|---|
| Keyword: | Maybank's , Markov Chain , Arma Model |
| Subjects: | Q Science > Q Science (General) |
| Department: | SCHOOL > School of Science and Technology |
| Depositing User: | MDM FAUZIAH MATSIN |
| Date Deposited: | 13 May 2013 13:23 |
| Last Modified: | 13 Oct 2017 16:59 |
| URI: | https://eprints.ums.edu.my/id/eprint/5901 |
Actions (login required)
 |
View Item |