In this paper, we apply textual analysis and machine learning algorithms to construct an index capturing trade tensions between US and China. Our indicator matches well-known events in the US-China trade dispute and is exogenous to the developments on global financial markets. By means of local...