A Comparative Study of Penalized Regression Methods in Estimating and Predicting Economic Growth in Nigeria

Abstract

This study investigated macroeconomic indicators-internal debt (INDT), external debt (EXDT), interest rate (RINR), exchange rate (REXR), and economy openness (OPEN) in modeling, estimating, and predicting Nigeria's economic growth (RGDP). To address multicollinearity and outlier problems, penalized least squares regression methods (LASSO, ridge, and elastic net) were employed. Results from the LASSO model indicated that INDT, RINR, REXR, and OPEN positively influenced RGDP, which contributed 4.27%, 0.40%, 0.49%, and 0.52%, respectively, whereas the influence of EXDT was negative (−0.97%). Ridge and elastic net estimations supported these results, with slight variations in coefficient magnitudes. All models emphasized the adverse effect of EXDT on RGDP. The evaluation and predictive power of the model metrics revealed that LASSO outperformed the other methods, revealing minimum root mean square error (0.2895), mean absolute error (0.2174), and mean absolute percentage error (2.13%), while explaining 91.9% variations in RGDP. Consequently, the results affirmed LASSO penalized technique as the most efficient variable selection with stable prediction under correlated and highly extreme dataset. Overall, the findings highlighted the crucial roles of internal borrowing and economy openness as growth-driven predictors, while stressing the adverse implications of reliance on foreign loans. Hence, the government should ensure proper management of the internally borrowed funds and channeled such funds by investing heavily in infrastructure, technology advancement, and innovation systems to maximize the growth benefits of the loans; application of penalized regression methods, particularly LASSO, should be mainstreamed into economic forecasting units within the government and research institutions to improve evidence-based policy formulation.

Received: 13 June 2025 | Revised: 4 January 2026 | Accepted: 23 January 2026

Conflicts of Interest

The author declares that he has no conflicts of interest in this work.

Data Availability Statement

The data that support the findings of this study are openly available in the central bank of Nigeria statistical bulletin at https://www.cbn.gov.ng.

Author Contribution Statement

Ayooluwade Ebiwonjumi: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration.