Impact of financial development and internet use on export growth: New evidence from machine learning models

The current study empirically examines the impact of financial sector development and Internet use on exports value for thirty Chinese provinces from 2000 to 2018 using the Panel-Corrected Standard Error (PCSE) estimation method and Gaussian Process Regression (GPR) machine-learning model. The PCSE method shows that (i) the Internet use increases China's exports; (ii) the impact of the Internet on export growth is larger in high-middle developed provinces; (iii) Internet use increases exports in high-middle developed provinces; however, it has less contribution to exports in low-developing China's provinces; (iv) financial development does not influence export value in the three panels. We also find that the GPR machine-learning model is more robust in predicting the exports growth in China based on the financial development and Internet use, which shows that population, Internet use, GDP, and financial development are the most important factors to predict exports growth in China. [Display omitted] •We estimate and predict the impact of financial development (FD) and internet use on export value.•Internet use increases China's exports in all the Chinese provinces•Impact of Internet on export growth is larger in high-middle developed provinces•Financial development does not influence export value in the three considered panels.•The machine-learning model show that Internet and FD are key factors in predicting exports growth.