Page 26 - Azerbaijan State University of Economics
P. 26

Murad Y. Yusıfov: Econometrıc Assessment Of Optımal Interest Burden: Case Study For Azerbaıjan

                    The study was based on a polynomial regression assessment in which the effects of

                    the interest burden on bank profits in Azerbaijan. The dummy variable of COVID-19

                    used in the polynomial model takes into account the lockdown effect. Let us now
                    consider the econometric assessment of the impact of the interest burden on the profits
                    of the banking sector for the case study of country (Table 2):

                    Table  2.  Polynomial  Regression  Assessment  of  the  Relationship  between  the
                    Profit of the Banking Sector and the Interest Burden in Azerbaijan
                            Variable                              Dependent variable:
                                                            Profit of The Banking Sector**
                                                                        Azerbaijan
                        c                                                -535,6153
                                                                       (-2.547774)*
                                                                         60207,95
                          
                                                                        (2.633506)
                         2
                                                                         -1482026
                          
                                                                       (-2.424778)*
                                   ∗                                     -15239,3
                                     
                                                                        (-2.404782)
                                   2
                                   ∗                                     700811,6
                                     
                                                                        (2.272575)*
                        AR(2)                                                -
                                                                             -
                        AR(12)                                           -0.195670
                                                                       (-1.743196)*
                        MA(1)                                            -0.947746
                                                                       (-15.74095)*
                        MA(6)                                                -
                                                                             -

                     R-squared                                           0.357674
                     Observations (months)                                  48
                     Observations (years)
                     * t - statistics are shown in bracket.
                     The  notes  **,  represent  the  stationarity  significance.  Test  shows  that,  the  first
                     differences  of  all  logarithmic  variables  are  stationary  at  the  1%,  5%  and  10%
                     significance levels. Augmented Dickey-Fuller (ADF) procedure has been realized
                     in Eviews (Annex 2). While evaluating of the parameters of model, it is defined that
                     the errors follow the normal distribution.
                    Source:  Estimation  made  via  EViews  software  and  summary  table  of  results
                    compiled by the author.




                                                           26
   21   22   23   24   25   26   27   28   29   30   31