Engagement in Earnings Conference Calls: A Multi-Method Examination

• In Event: Picks of the 2019 ABO Coordinators

Abstract

Research on earnings conference calls documents that the question and answer (Q&A) portion is more informative than the presentation portion, noting the role of analyst involvement. However, prior studies focus on the behavior of managers and analysts individually in earnings calls. In this study, we use the interaction itself as our unit of analysis, and examine whether characteristics of the interaction provide information to market participants. Drawing on research in psychology and sociolinguistics, we provide experimental and archival evidence on the determinants and consequences of conversational engagement between managers and analysts in earnings calls. We use linguistic style matching (LSM), a form of verbal coordination, as our proxy for conversational engagement. Prior literature suggests conversational engagement can be a signal of the importance of the interaction, which may be informative to market participants. Using a controlled experiment with MBA students, we first demonstrate that LSM is a reasonable proxy for conversational engagement between managers and analysts. That is, LSM is greater when managers have more incentives to engage with analysts. Next, using a unique hand-collected archival dataset comprised of audio recordings and textual transcripts from over 2,400 earnings calls, we provide convergent evidence that LSM is associated with manager and analyst incentives for engagement. Finally, mapping these calls to concurrent intra-day stock price movements, we show that LSM is informative to market participants. Our findings contribute to the literature on the information content of earnings calls.

Authors

• Kristina Rennekamp, Cornell University

• Mani Sethuraman, Cornell University

• Blake Steenhoven, Cornell University