Forecast of Belex15 and Belexline movement using ARIMA model

Abstract

Financial markets are known to be volatile and often unpredictable. Movement of stock indices values and stock prices over time can be labelled with big oscillations during time, so financial institutions and other investors are constantly working on forecasting these movements, in order to adjust their business decisions and increase their profit. Aim of this research is to determine whether ARIMA model, which is often used for short-term forecast, is suitable for forecasting of Belgrade Stock Exchange indices movement. Subject of this research are Belex15 and Belexline indices and based on daily data from 5^th January 2009 until 31^st March 2022, ARIMA model was used to forecast indice values for following 11 trading days (from 1^st April until 15^th April 2022). Methodology of this research contains of presenting and analysing empirical data of Belex15 and Belexline movements, choosing suitable ARIMA model for forecast and forecasting indice values, which were later compared with real indice values. Obtained results point to high accuracy of forecasted values and lead to the conclusion that ARIMA model is corresponding econometric method for short-term forecast of Belgrade Stock Exchange indices.