aDepartment of Financial Management, Science and Research Branch, Islamic Azad University, Yazd, Iran
bDepartment of Management, Yazd Branch, Islamic Azad University, Yazd, Iran
cDepartment of Economics, Abarkouh Branch, Islamic Azad University, Yazd, Iran
Online published on 27 March, 2014.
The present study is aimed to forecast the stock price index as the main tool for forecasting in stock exchange trend. The results can be useful for investors and decision-makers to reduce the risk of their investment. The data analyses showed forecasting capability of stock price index in ARIMA (0,1,1). However, the fast changing financial market hampers the forecasts to achieve the required data as the current qualitative models such as ARIMA suffer from the lack of enough data from the past. Fuzzy forecasting models suit for situations with lack of enough data. Combining different models or using different models at the same time is one solution to cover the shortcomings of one model. Thus, the present study uses classic accumulated auto-regression changing mean models to deal with the lack of data and to gain more accurate forecast of time series, these models were used in combination with fuzzy logic (FARIMA). Taking into consideration the research questions, the future trend of stock price in 2011 was forecasted and the results by ARIMA and FARIMA model were compared. Our findings showed better performance of the proposed combined model for forecasting stock price in Tehran Stock Exchange (TSE).
Forecasting, price index of Tehran Stock Exchange, time series, ARIMA model
