Buzzwords?

Abstract

Identifying emerging technologies is challenging since the frontier of new technologies is always rapidly evolving and given the potential for significant uncertainty and hype surrounding these technologies. Using the pre-trained Bidirectional Encoder Representations from Transformers (BERT), a deep learning model for textual analysis, we construct a dictionary of terms and phrases that identifies emerging technologies as they material over time. We then study the response to the discussion of emerging technologies by analyzing the earnings conference calls of U.S. public firms from 2007 to 2019. We find that the discussion of emerging technologies is associated with a significantly positive immediate stock market reaction. This reaction is more pronounced when the discussed technologies are in their nascent stages, and when firms have prior innovation experience in related technology domains. Further, there is no evidence of a negative post--earnings call drift for the firms that discuss emerging technologies in their earnings calls. Our results suggest that the discussion of emerging technologies in earnings calls is not just hype, but conveys credible information to investors.

Keywords: Deep Learning, Natural Language Processing, Buzzwords, Emerging Technologies, Hype, Earnings Conference Calls

JEL Classification: O30, O31, O32, G14