Function simulation of FDI, foreign trade and regional GDP in China

Purpose – The purpose of this paper is to simulate the function of foreign trade, foreign direct investment (FDI) and regional gross domestic product (GDP) in China, explore how these two variables affect regional GDP together and provide evidence to export‐led growth (ELG) and FDI‐led growth. Design/methodology/approach – Artificial neural network (ANN) is introduced in the model. This nonlinear and adaptive computation obtains a three‐dimension function that is different from linear models. Findings – New evidence was found for ELG and FDI‐led growth with data of 28 regions in China in the period of 1994‐2005. The simulation reveals that with foreign trade and FDI scale varying, marginal GDP in different Chinese regions is positive. Because of the nonlinear system, a wave pattern of marginal GDP was found and an optimal scale of foreign trade and FDI for Chinese regions. Results in the simulation also indicate the possibility of economic deconcentration in some Chinese regions. Originality/value – New evidence is provided for ELG and FDI‐led growth. Different from conventional methods, ANN model as a nonlinear system is introduced in the study in which optimal scale of foreign trade and FDI for Chinese regions is obtained.