# Evaluation of the Impact of Selected Financial Indicators on Foreign Direct Investment in Bangladesh: A Nonlinear Modeling Approach

## TL;DR

This study explores how economic indicators affect foreign direct investment in Bangladesh using a nonlinear model that outperforms traditional methods.

## Contribution

The novelty lies in using a generalized additive model to capture nonlinear relationships in FDI prediction for Bangladesh.

## Key findings

- FDI in Bangladesh is significantly linked to GDP, trade openness, and other economic indicators in a nonlinear fashion.
- The generalized additive model (GAM) outperformed linear and polynomial regression in predicting FDI with high accuracy (R2 = 0.987).
- Nonlinear patterns between FDI and economic factors suggest the need for advanced modeling approaches like GAM for better predictions.

## Abstract

Background: Foreign direct investment (FDI) is a steadfast contributor to capital flows and plays an indispensable role in driving economic advancement and emerging as a pivotal avenue for financing growth in Bangladesh. Therefore, this study identifies the factors that influence FDI inflows in Bangladesh. Moreover, the authors explored the more appropriate model for predicting FDI by comparing the efficacy of other models' predictions.

Methods: This study is based on secondary data over the period 1973 to 2021 and collected from the publicly accessible website of the World Bank. A generalized additive model (GAM) was implemented for describing the proper splines. The model's performance was assessed using the modified R-squared, the Bayesian information criterion (BIC), and the Akaike information criterion (AIC).

Results: Findings depict a significant nonlinear relationship between Bangladesh's FDI and key economic indicators, including GDP, trade openness, external debt, gross capital formation, gross national income (GNI) and government rates of exchange, total reserves, and total natural resource rent. It is also observed that the GAM (R2 = 0.987, AIC = 608.03, and BIC = 658.28) outperforms multiple linear regressions and polynomial regression in predicting FDI, emphasizing the superiority of GAM in capturing complex relationships and improving predictive accuracy.

Conclusion: A nonlinear relationship is observed between FDI along with the covariates considered in this study. The authors believed that this study's findings would assist in taking efficient initiatives for FDI management and proactive economic indicator optimization to empower Bangladesh's economic resilience and foster sustainable growth. The analysis revealed that FDI and its related risk factors follow a nonlinear pattern. The study recommends using the GAM regression as a reliable method for predicting FDI in Bangladesh. The authors suggest that the findings can guide policymakers in developing strategies to increase FDI inflows, stimulate economic growth, and ensure sustainable economic development in Bangladesh.