Oh What a Beautiful Morning!
The Time of Day Effect on the Tone and Market Impact of Conference Calls
Jing Chen
Elizabeth Demers
Baruch Lev
July 2013
Abstract
Using textual analysis software, we examine whether and how the tone of the question and answer (“Q&A”) portion of earnings-related conference calls varies with the time of day. We find that the tone of the conversations between analysts and managers becomes increasingly negative with time, which we conjecture to be due to accumulating mental and physical fatigue over the course of the day. Textual uncertainty exhibits a similar pattern, with conversational tone becoming more wavering and less resolute later in the day. We document that conversational tone has economic consequences; more negatively toned conversations are associated with more negative abnormal stock returns during the call period and immediately thereafter. Notwithstanding the negativity associated with later day calls, firms exhibit significant “stickiness” in their choice of call time; having initiated the earnings conference call in the afternoon in the prior quarter is the most significant determinant of their doing so in the current quarter, dominating the sign of the earnings news and alternative measures of the firm’s need for equity capital. Analysis of post-call (50 days) returns indicates that there is an initial negative overreaction to bad news earnings information and, incrementally, to calls initiated in the afternoon, that eventually reverses. In contrast, the negative impact of tone deterioration on stock returns, documented here, does not reverse. To the best of our knowledge, this is the first study to document the effects of human physiological and mental factors on corporate communications with investors.
Introduction
Conference calls have become the major channel of corporate communications to outside stakeholders (Frankel, Johnson & Skinner (1999); Bushee, Matsumoto & Miller (2003); Skinner (2003)). Earnings related conference calls are typically conducted within a few hours to a day following the quarterly earnings release, and consist of a managerial presentation (that generally reiterates the main content of the earnings press release), followed by a question-and-answer (Q&A) session with analysts and investors. As a result of the analyst-management interaction, the Q&A portion of the call elicits significant incremental information over the earnings announcement and managerial presentation (Matsumoto, Pronk & Roelofsen (2011)). Various aspects of conference calls have been examined by researchers, such as the impact of the call on stock prices and volumes, or the effect of the call’s tone and other characteristics on prices and analyst recommendations. In the current study, we examine a new dimension of conference calls: whether and how the tone of call conversation and its consequences varies with the time of day, reflecting the mood and the levels of mental and physical state of call participants. Unlike corporate press releases and the text of mandatory filings that have been examined in prior financial linguistics studies, the Q&A portion of the conference call is not scripted by lawyers and communications experts, and is more likely to involve natural, less inhibited use of language. We therefore expect that our linguistic measures will capture the spontaneous tone and nuance of the communications between parties to the call.
Using multiple proxies for two linguistic dimensions that we term negativity and textual uncertainty, we find that the time of day at which the call was initiated (our proxy for call participant mental and physiological state), impacts in a remarkably systematic manner these tonal aspects of manager-analyst interchanges. Specifically, as the day wears off, participants’ tone gets increasingly negative and uncertain (wavering). This changing conversational tone has real economic consequences: our measures of linguistic tone are associated with the firm’s intraday call period stock returns and return volatilities, as well as the 50-day post-call drift returns. To the best of our knowledge, ours is the first study that relates the tonal aspects of financial communications to the time of day.
