Page 12 - Azerbaijan State University of Economics

THE JOURNAL OF ECONOMIC SCIENCES: THEORY AND PRACTICE, V.81, # 1, 2024, pp. 4-21

Table number (04) shows that the variables are not significant at the 5% level, meaning
they are non-stationary at level due to the presence of a unit root, which means
accepting the null hypothesis. After re-testing the variables at the first difference, both
series were stable at the 1% significance level, i.e., free from the unit root, leading to
the acceptance of the alternative hypothesis and the rejection of the null hypothesis.
Therefore, the study variables are integrated of order one I (1).

Results of Estimating the Study Model
To understand the impact of FDI on economic development, the model is estimated, and
various diagnostic tests are conducted to ensure the authenticity of the estimated results .

Estimation of the FARDL Model
The stability study results reveal that the study variables, economic development (GDP)
and FDI (FDI), are integrated of the same order I(1), suggesting the possibility of a long-
term cointegration link between these variables within the FARDL methodology
framework, which is considered one of the best models for studying cointegration .
Choosing the Optimal Lag Length for the FARDL Model
Before estimating the model, it's crucial to determine the optimal lag periods. Based
on the Akaike Information Criterion (AIC), the FARDL (4,2) model was chosen as it
provides the lowest value for this criterion among a set of models as shown in the
following figure:
Figure 2: Optimal Lag Length for the ARDL Model

Source: Compiled by the researcher depending on: outputs of EViews 12.

12