Vector Auto Regression Case Study

The methodology of this study is use Augmented Dickey Fuller (ADF) test statistic to determine whether the variables had been used are stationary or non-stationary. Vector Auto Regression (VAR) method is apply in this study. The advantages of VAR is time series can be exhibited at the same time. The VAR methodology is revises for autocorrelation and endogeneity parametrically using vector error correction model (VECM) specification. Base on Johansen (1988; 1995), the benefit of VECM is that it prevents the bias that takes place in OLS estimates of cointegration relations when use the Engle-Granger procedure. The procedure of Johansen econometric methodology has develops. The starting point in VAR of order m is:
Zt = U + A1z t-1 +…….. Am Zt-m
According to the Harris (1995), specifying the system this way will have information on the short-run and long-run adjustments to changes in Zt through ∏ and ri estimates. According to the Johansen (1988; 1995) the suggested two likelihood ratio test statistics for cointegration and these are trace (λtrace) and maximum eigenvalue (λmax).

Conclusion
Empirical finding Unit Root Test
The first step before estimate the regression equation is using Unit Root Test by using the Augmented Dickey Fuller test statistic. The results of the Unit Root Test is shown that all the variables, excluding balance of payment and domestic credit, are non-stationary in levels, but stationary in first difference. Cointegration Test and VECM Results
Cointegration Test is the next step after establishing the Unit Root Test. Johansen Cointegration test have been applied and it is using two statistics, which are trace ( λtrace )and maximum eigenvalues (λmax). The diagnostic statistics were executed on the unlimited VAR and the VECM and the results show that residuals are multivariate normal, no Heteroscedasticity and no serial correlation.
The cointegration relation the balance of payments equation in long-run is showing below. The results of the forecast long-run balance of payments equation are presented in below and the t-statistics are in
When the fiscal balance increase 1%, the BOP increases by 0.13%. In Namibia, government revenue is obtain from SACU receipts and transfer has substantial implications on the balance of payment movements. The coefficient of exchange rate is negative and it found to be highly statistically insignificant. The impact of the nominal exchange rate and the price differentials between domestic and foreign prices does not strongly impact on the balance of payments. The instant impact of the exchange rate is felt on the capital and current account but its impact can only become significant over a lagged period. The coefficient on GDP is positive and statistically significant. When GDP increase 1%, the balance of payments increase by 6.38 percent. The export expansion when there are positive impact of GDP. When the export increase, the trade balance and balance of payments increase and the relationship is positive. When interest rates increase 1%, it causes the balance of payments to increase by 3.2 %. This is consistent with theoretical expectations that a positive interest rate would attract capital inflow and increase the balance of