Page 77 - Azerbaijan State University of Economics

Murad Yusifov: Modelling the inflationary processes and forecasting:an application of ARIMA,
SARIMA models

Coefficients value of the model within 95% confidence interval were given in the

Annex1.(See.Annex.1.Table.9.)

Table.10. ARIMA model emprical results

ARIMA SARIMA Model ARIMA SARIMA
Dövr Model CPI CPI Actual CPI Deviation(+/-) Deviation(+/-)
2014M07 99,47 99,47 99,21 0,3 0,3
2014M08 100,03 100,03 99,99 0,0 0,0
2014M09 100,41 100,21 101,10 -0,7 -0,9
2014M10 100,61 100,44 100,25 0,4 0,2
2014M11 100,67 100,48 100,10 0,6 0,4
2014M12 100,63 101,11 100,50 0,1 0,6
2015M01 100,54 99,97
2015M02 100,44 100,29

2015M03 100,35 100,28
2015M04 100,29 100,22
2015M05 100,26 99,87
2015M06 100,24 99,80
2015M07 100,24 100,07
2015M08 100,25 100,23
2015M09 100,27 100,30
2015M10 100,28 100,34
2015M11 100,29 100,35
2015M12 100,30 100,34

4. Conclusion

Modelling the inflationary processes and forecasting the estimates are of great

importance while developping the investment projects, indexation of wages, predetermining the

macroeconomic policy and preventive measures. The advanced forecasting models such as

SARIMA and ARIMA have been applied and achieved results in this study. For forecasting the

inflation the empirical results have been obtained on the seasonal avtoregressive integrated

moving average model SARIMA(1,0,1) (0,0,1)x12 and non-seasonal avtoregressive integrated

77