Scientific Beta

This paper analyses a firm-level measure of exposure to shifts in trade policy and documents it captures substantial heterogeneity across firms in their reaction to trade policy shifts. This measure combines multiple dimensions of trade policy exposure and exploits information in stock return covariance in addition to industry and firm characteristics. We show in detailed validation tests that the measure captures heterogeneous exposure of firms to trade policy shocks in terms of both stock price reactions and sentiment on trade policy.

Trade policy influences the financial performance of firms. Moreover, shifts in trade policy will have heterogeneous effects across such businesses; some may benefit from a more globalised market via reduced input costs or more export opportunities, while others may benefit from the introduction of protectionist policies via a reduction in foreign competition. A measure that captures the impact of shifts in trade policy on firms is therefore a useful tool for researchers in finance or economics and financial market participants alike.

We propose such a firm-level measure of exposure to trade policy shifts. The measure draws on data regarding the tradability of produced goods, the share of output exported, and corporate risk disclosures. Furthermore, it also incorporates information from stock returns. This multi-dimensional measure has the advantage that it uses a broad set of information. International trade is complex, so we cannot expect a single data source to capture exposure to shifts in trade policy accurately. Combining multiple metrics also reduces the impact of noise present in each of them individually.

We show that the measure reliably captures out-of-sample differences in exposure to trade policy shifts across firms. First, we find that firms with a high exposure according to our measure have a more negative stock price reaction around tariff announcements than firms with a low exposure. This result is both statistically and economically significant, with a 140bp difference during the week following the announcement. Second, in times of rising trade tensions, high exposure firms discuss trade policy with a more negative sentiment than low exposure firms in their quarterly earnings calls. Supportive out-of-sample evidence based on two entirely different data sources strengthens our confidence in the validity of the measure. Moreover, we show that standard firm classifications by size and industry are insufficient to capture such heterogeneity in trade policy exposures and we present evidence both in US and international data.

While the focus of this paper is on the measure of exposure to trade policy shifts and its validation, we also discuss a variety of potential applications. The firm-level granularity and reliance on ex-ante available data makes the measure suitable for a wide range of applications. Among others, we expect the measure to be useful for researchers interested in the economic effects of international trade policy or asset pricing, or for investors who want to manage the trade policy risk exposure in their equity portfolio.