Working Paper: CEPR ID: DP14573
Authors: Tarek Hassan; Stephan Hollander; Laurence van Lent; Markus Schwedeler; Ahmed Tahoun
Abstract: We introduce a new word pattern-based method to automatically classify firms' primary concerns related to the spread of epidemic diseases raised in their quarterly earnings conference calls. We construct text-based measures of the costs, benefits, and risks listed firms in the US and over 80 other countries associate with the spread of Covid-19 and other epidemic diseases. We identify which firms and sectors expect to lose/gain from a given epidemic and which are most affected by the associated uncertainty. Our new automatic pattern-based method shows how firms' primary concerns (varying from the collapse in demand and disruptions in their production facilities or supply chain, to financing concerns) are changing over time and varying geographically as epidemics spread regionally and globally. We find that the Covid-crisis manifests itself at the firm-level as a simultaneous shock to both demand and supply. In prior epidemics, in contrast, firm discussions center more on shortfalls in demand. In 2020, supply and financing-related concerns are relatively more salient in regions where the spread of Covid-19 is less contained.
Keywords: epidemic diseases; pandemic; exposure; virus; firms; uncertainty; sentiment; machine learning
JEL Codes: I15; I18; D22; G15; E0; F0
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| COVID-19 mentions (I12) | firm concerns (demand vs. supply) (D22) |
| COVID-19 exposure (I14) | stock returns (G12) |
| firm concerns (demand vs. supply) (D22) | stock returns (G12) |
| time of discussion (C41) | tone of discussions (Z00) |
| sector (R38) | firm concerns (G32) |