By using an approach based on text data and supervised machine learning, this paper propose a novel measure of investment plans in the firm-level. I combine the procedure of Han, He, Rapach, and Zhou (2020) with the idea of flexible dictionary of Lima, Godeiro, and Mohsin (2020) and, I test a novel measure of investment plans based on text data from Management Discussion and Analysis (MD&A) disclosure in 10-K filings. In this study, the sample includes all US publicly traded firms in the period between January 1995 and December 2019. I build a unique dataset by merging information from multiple data sources. The annual firm-level financial and accounting data, I obtain from Compustat. The firms’ 10-K filings are from the SEC Edgar database and the monthly US stock returns from the Center for Research in Security Prices (CRSP). The main find is that the words of MD&A matter to predict firm fundamentals, even in the case of investment growth, which is empirically challenging to measure in the firm-level.ddddBy using an approach based on text data and supempirically challenging to measure in the firm-level.
Comissão Organizadora
Anderson Odias da Silva
Claudia Yoshinaga
Ricardo D. Brito
Felipe Saraiva Iachan
Vinicius Augusto Brunassi Silva
Complete tool for submission, evaluation and publication of proceedings for scientific events.
