Combined Application of State Space in ARIMA Form Model and Monte Carlo Simulation Method to Forecast TEPIX Index

Subject Areas : Financial Knowledge of Securities Analysis

Aghigh Farhadi ¹ , Farhad Ghaffari ²

1 - دانش آموخته کارشناسی ارشد رشته علوم اقتصادی ، دانشکده مدیریت و اقتصاد ، دانشگاه آزاد اسلامی ، واحد علوم تحقیقات تهران
2 - استادیار ، دانشکده مدیریت و اقتصاد ، دانشگاه آزاد اسلامی ، واحد علوم تحقیقات تهران

Received: 2014-11-29 Accepted : 2015-01-10 Published : 2015-06-22

Keywords: State Space, Monte Carlo simulation, Forecasting Accuracy, Efficiency,

Abstract :

In this study, we estimated the parameters using the State Space model described inARIMA form. We’ve also used the Monte Carlo Method for simulating the process in10000 reputations. Then the estimated parameters and the Monte Carlo simulationmethod are used to forecast TEPIX index, including 739 observations as an in-sampledata from 21th of January 2011 to 19th February 2014 and 59 observations from 20thFebruary 2014 to 21th May 2014 as an out of sample data . Furthermore, For moreinvestigation we’ve considered different horizons of forecasting, short-term (equal to 1week), mid-term (equal to 1 month) and long term (equal to 3 month). The results showedthat Tehran stock market data has enough efficiency to forecast them, and showed that theState Space in Form ARIMA model and the Monte Carlo simulation method can be usedas a predictive algorithm for TEPIX index and other indices with similar nature.