A0259
Title: Actual events vs. actual reporting: Modeling firm performance under environmental uncertainty using machine learning
Authors: Minh Nguyen - University of Hawaii at Manoa (United States) [presenting]
Abstract: Not all companies respond the same to natural disaster events. Two ways that natural disasters affect firm performance are studied: actual events vs. actual reporting. We consider the billion-dollar natural disasters in the United States as the actual events and the number of words related to natural disasters in the Management Discussion and Analysis section in Form 10-Ks filing by the US public companies as the actual reporting. The aim is also to compare the performances of classification and regression trees (CART) and neural networks with the benchmark model-based linear regression model in predicting the performance of U.S. public companies under environmental uncertainty. We find that both actual events and actual reporting of natural disasters in year $t$ negatively affect the return on assets (ROA) in year $t+1$. Also, the actual natural disasters in year $t$ negatively affect sales growth in year $t+1$. Moreover, we find that the environmental uncertainty variables are much less important than the traditional financial statement variables in predicting firm performance with the CART model. Comparing CART, neural networks, and linear regression models, we find that CART and neural networks outperform linear regression models in predicting firm performance.
