I am trying to reproduce SPSS results with python.
I'm using statsmodels SARIMAX package in python.
My model produces forecasting results which shows sudden drop at '2020-08-01'.
After '2020-08-01', the result shows consistent pattern.
Why does my model forecast sudden drop at '2020-08-01'?
This is the summary of my model.
And this is the last few observations in my data.
2018-11-01 67.238591
2018-12-01 69.397353
2019-01-01 73.369985
2019-02-01 80.079167
2019-03-01 86.074545
2019-04-01 76.248803
2019-05-01 66.693503
2019-06-01 61.937356
2019-07-01 53.014063
2019-08-01 47.497356
2019-09-01 44.001648
2019-10-01 37.382052
2019-11-01 31.786453
2019-12-01 25.112575
2020-01-01 19.502593
2020-02-01 15.443444
2020-03-01 12.736181
2020-04-01 12.567810
2020-05-01 12.090781
2020-06-01 7.166829
I guess it's because my model is ARIMA(1,0,2) which uses two MA terms.
Am I right?
Could you interpret this results with manual calculation or mathematical formula?
