Search results

This study aims to find the impact of the trade war between the USA and China on Asian economies. Apart from macroeconomic variables associated with trade, this study explicitly creates a trade war scenario and trade war participant dummies. Using the neural network multilayer perceptron, this study checks for the causal linkages between the predictors and target output for the panel of Asian economies and the USA.

A conceptual model of the after effects of trade war in a quadrant is developed. Variables related to trade and tariffs are included in the study for a panel of 19 Asian economies. The feedforward structure of neural network analysis is used to identify strong and weak predictors of trade war.

The hidden layers of the multilayer perceptron reveal the inconsistency in linkages for the predictors’ services exports, tariff measures, anti-dumping measures, trade war scenario dummy with gross domestic product. The findings suggest that to curtail the impact of the trade war on Asian economies, predictors with neural evidence must be paid due weightage in policy determination and trade agreements.

The study applies a novel and little explored AI/ML technique of Neural Network analysis with training of 70% observations. The paper will provide opportunity for other researchers to explore techniques of AI/ML in trade studies.