การศึกษาเปรียบเทียบการพยากรณ์ราคาหลักทรัพย์กลุ่ม SET100 ของตลาดหลักทรัพย์แห่งประเทศไทยโดยวิธีมอนติคาโลและวิธีบ๊อกซ์-เจนกินส์

Abstract

A purpose of this research is to analyze the risk of stock and study accuracy of Monte Carlo method and Box-Jenkins method in forecasting the stock market prices in Thailand which is SET100 during July 2005 to December 2009. The accuracy was tested by Mean Square Error (MSE) with secondary data which is a time series data of the 50 closing stock market prices. According to analyzing the risk of 47 stocks, the results of the study can be divided into three groups. The signs of risk coefficient are plus, minus, and both plus and minus. This research found that there are 18, 10, and 19 stocks with plus, minus and both signs, respectively. Examining the accuracy of stock market prices forecasting showed that Monte Carlo method can forecast 27 stocks correctly which accounted for 54%. Box-Jenkins method can predict 23 stocks which accounted for 46%. For examining the accuracy of stock market prices forecasting in each type of business, Box-Jenkins method is accurate in 4 types of stock market which accounted for 57.14% and Monte Carlo method can forecast accurately 2 which accounted for 28.57%, the rest of type of stock market both method can identical forecasting which accounted for 14.29%. The conclusion of SET100 testing hypothesis showed that Monte Carlo method is better than Box-Jenkins method in forecasting; therefore, Monte Carlo method can be used to forecast in that period accurately.