C. Model Testing
The study is concerned about the effect of factors that
affect the stability of palm oil used Auto Regressive (AR)
Integrated (I) that is Statistic Model accompany with
independent variables. The results are as follow .
1) Augmented Dickey-Fuller Test
The data used in this research are fixed and can be
considered in return to the 11variables. See fig. 5
Fig. 5 shows as the ADF Test Statistic level of all variables
which contain Unit Root or time series is qualified Non
Stationary. This means that all variables have stationary qualifications at First Differencing include FFB_F Price,
CPO_T Price, CPO_M Price, CPO_L Price, B100 Price,
RPO_T Price, RPO_M Price, Cooking Oil Price,
GDP_Percapita, SBO_W Price, RSS_F Price and QFFB_F.
The value from the test was conducted by considering the tautest value is greater than the tau-critical. The researcher takes
the first difference to calculate the equilibrium relationship
and found that variables have long-term equilibrium
relationships. The researcher uses that suitable variable to
build ARIMAX Model.