MODELIRANJE I PROGNOZIRANJE BDP BOSNE I HERCEGOVINE PRIMJENOM ARIMA MODELA
MODELLING AND FORECASTING GDP IN BOSNIA AND HERZEGOVINA USING ARIMA MODELS
Author(s): Ademir Abdić, Emina Resić, Adem Abdić, Fahir KanlićSubject(s): National Economy, Supranational / Global Economy, Methodology and research technology, Financial Markets
Published by: Ekonomski fakultet u Sarajevu
Keywords: quarterly GDP; ARIMA models; unbiasedness; accuracy of forecasting;
Summary/Abstract: In the most EU countries the first estimate of GDP is available 30 days after the end of the quarter to which the assessment relates (Eurostat, 2017). Possibilities of using Box-Jenkins methodology for making short-term forecasts of BiH GDP have been explored in this paper. The variable GDP BiH is available for period 2000q1-2016q4. Due to fact that a priori is very difficult to rank the different models of GDP forecast, their relative rank will be assessed based on empirical data. The final choice of the model for forecasting quarterly B&H GDP was selected on the basis of a comparative analysis of the predictive efficiency of the analyzed models for in-sample-period (2000q1-2014q4) and for a quasi-out-of-sample period (2013q1-2016q4). Based on the obtained research results, the seasonal ARIMA(0,0,0)(1,1,0)4 model has been selected as the most efficient model in the forecasts of quarterly B&H GDP.
Journal: Zbornik radova - Sarajevo Business and Economics Review (SBER)
- Issue Year: 2019
- Issue No: 37
- Page Range: 9-34
- Page Count: 26
- Language: Bosnian