The macroeconomic impact of trade policy : a new identification approach

This paper examines the effects of trade policy shocks on the US economy using a novel identification strategy that combines narrative information with stock market data. We construct a new dataset of daily trade policy statements from 2007 to 2019, enabling us to capture a broad range of policy actions. By analyzing stock price reactions of trade-exposed and non-trade-exposed firms around these statements, we identify unanticipated trade policy shocks. Using the local projections method, we assess asymmetries and non-linearities based on the sign and size of shocks. We find that the economic effects of trade liberalizations and protectionism are symmetric, with no evidence of non-linearities. However, foreign-initiated trade shocks have a larger impact than US-initiated ones, and policy implementations have a stronger effect than announcements alone. Finally, we explore whether relying on President Trump's trade-related tweets, rather than official statements, alters the estimated effects.