Abstract
This study investigates text as treatment mining from Form 10-Ks to quantify causal effects of perceived natural hazards on firm profitability. To do that we employ natural hazards data from the National Oceanic and Atmospheric Administration (NOAA) and perceived natural hazards exploited from Form 10-Ks of 14,994 US public companies. Our final sample includes 148,603 firm-year observations. We find that perceived natural hazards negatively affect firms' return on assets (ROA). This result is robust and consistent between models using linear regression, double machine learning (DML), and meta-learners. The key implication from our work is that firms are not only affected directly by natural hazards themselves as shown in the existing literature but also the perception of them.
Recommended Citation
Nguyen, Minh; Yeh, Ting-Tsen; Xiao, Yuanzhang; and Daniel, Shirley J., "Text as treatment from Form 10-K: Causal effects of perceived natural hazards on firm profitability" (2023). International Conference on Information Systems 2023 Special Interest Group on Big Data Proceedings. 4.
https://aisel.aisnet.org/sigbd2023/4