Global Journal of Human Social Science, E: Economics, Volume 21 Issue 4

The data was not stationary. c) Differenced Unit Root Series Prob. Lag Max Lag Obs D(F.D) 0.02 12 14 226 D(D.D) 0.00 13 14 227 D(E.G) 0.00 12 14 226 Method Statistic Prob.** ADF - Fisher Chi-square 180.14 0.0000 ADF - Choi Z-stat -11.51 0.0000 The tested variables had no unit root this was later achieved on differencing. d) Correlation Analysis D.F.D D.D.D D.E.G D.F.D 1.000000 D.D.D -0.27 1.000000 D.E.G 0.17 0.53 1.000000 The correlation analysis revealed that both foreign debt and domestic debt are positively correlated with economic growth as evidenced by their coefficients 0.17 and 0.53, respectively. e) Johansen Cointegration Trace Test From Johansen test, the time series variables were not cointegrated. Consequently, the best model to use in the study was Vector Autoregressive Model (VAR). Vector Autoregressive Model (Var) Coefficient Std. Error t-Statistic Prob. C(1) 0.18 0.05 3.77 0.00 C(2) 0.13 0.05 2.78 0.01 C(3) -0.0002 0.00 -0.05 0.96 C(4) 0.55 0.07 7.98 0.00 C(5) 0.52 0.05 10.19 0.00 Determinant residual covariance 0.00 Equation: DEG = C(1)*DEG(-1) + C(2)*DEG(-2) + C(3) + C(4)*DFD + C(5)*DDD R-squared 0.64 Mean dependent var -0.01 Adjusted R-squared 0.63 S.D. dependent var 0.03 S.E. of regression 0.02 Sum squared resid 0.10 Durbin-Watson stat 1.34 Log Likelihood 587.84 The model is reliable for forecasting since Log Likelihood criteria {587.84>30} is an indication of the goodness of fit. With an R-Square of 64% the model is fit for prediction. This is interpreted that E.G can be explained by the variables in the model up to 64 % while the remaining 36% could be explained by other variables not in the current study. The model therefore was stated as; E.G = - 0.0001 + O.18E.G t-1 + 0.13E.G t – 2 + 0.5200 D.D + 0.5524 F.D f) Domestic Debt (D.D) and Economic Growth (E.G) The fourth objective of the study was to establish the effect of CBK overdraft and economic growth. The null hypothesis was therefore stated as follows; H 0 : Domestic Debt has no significant effect on Economic Growth. The table indicates that the regression weight of D.D and E.G was 0.5200 (p = 0.0000< 0.05) indicating existence of a positive and significant effect of D.D on E.G and hence the null hypothesis was rejected. This means that a unit increase in D.D causes E.G to expand by 0.5200. The amount borrowed should be matched with type of investment to ensure repayment is done on Volume XXI Issue IV Version I 46 ( E ) Global Journal of Human Social Science - Year 2021 © 2021 Global Journals Domestic or Foreign Debt ? A Choice of no Wrong Selection