Predictive Analysis of GDP-Debt Dynamics: An Econometric Approach

This study investigates the impact of external debt on economic growth in Nigeria over the period 2000 to 2023. Utilising a combination of econometric modelling, diagnostic testing, and predictive analytics, the research explores the debt- growth nexus, diagnoses econometric concerns, and forecasts future debt dynamics based on key macroeconomic indicators. The Ordinary Least Squares (OLS) regression results reveal that external debt exhibits a negative but statistically insignificant effect on economic growth, while variables such as foreign direct investment, government revenue, and inflation displayed limited explanatory power. Unit root tests indicate that most variables achieve stationarity after first or second differencing, except for inflation, which remains non-stationary despite several transformations and structural break adjustments. Diagnostic checks confirmed the absence of autocorrelation and heteroskedasticity; however, multicollinearity is detected among key predictors. This is addressed using ridge regression, which stabilises coefficient estimates and retains all explanatory variables for policy interpretation. Forecasting through the ARIMA (0,2,1) model projects a continued rise in external debt through 2028, while the Random Forest model identifies exchange rate, total debt, and government revenue as the most influential predictors. Marginal effects analysis further highlighted the significant roles of exchange rate and total debt in driving external borrowing. Scenario-based forecasting under alternative GDP growth rates shows minimal change in projected debt levels, suggesting that economic growth alone may be insufficient to reduce external debt burdens without comprehensive fiscal and structural reforms. The study concludes with actionable policy recommendations aimed at promoting sustainable debt management in Nigeria.