| 摘要: | It is believed that macroeconomic factors give great impact toward the stock market. The aim of this paper is to predict the FTSE Bursa Malaysia KLCI using macroeconomic variables with artificial neural network (ANN). The macroeconomic variables are considered in this study based on the Arbitrage Pricing Theory (APT) and other analytical models used in previous research. The macroeconomic variables used in this research are palm oil price, industrial production index, inflation rate, official reserve assets, exchange rate, and interest rate. This research also compares the efficiency of ANN model with other time series analysis using mean absolute percentage error (MAPE). The stock price is predicted in two different time periods: Period 1 (January 2002 to November 2015) and Period 2 (January 2010 to November 2015). The result shows that ANN approach has the lowest MAPE among all methods. MAPEs of ANN in shorter period are better than in the longer period. This is because the short period excludes the financial crisis impact in year 2008 and 2009. |
