Scripted Earnings Conference Calls as a Signal of Future Firm Performance

Transcription

1 Scripted Earnings Conference Calls as a Signal of Future Firm Performance Joshua Lee Olin Business School Washington University in St. Louis St. Louis, MO joshlee@wustl.edu January 2014 Abstract: I examine whether market participants infer negative information about future firm performance from managers scripted responses to questions received during earnings conference calls. I argue that firms script their Q&A session responses prior to periods of poor performance to avoid the inadvertent disclosure of information that can be used to build a lawsuit against the firm. Using a unique measure of conference call Q&A scripting, I provide evidence that scripted Q&A is negatively associated with future earnings and future cash flows, suggesting that, on average, firms script their Q&A when future performance is poor. I also find a negative market reaction to scripted Q&A and downward revisions in analysts forecasts following scripted Q&A, suggesting that investors interpret scripted Q&A as a negative signal of future firm performance. I also find that firms are less likely to guide future earnings when Q&A is scripted and that analysts forecasts are less accurate following scripted Q&A, suggesting that firms provide less information to market participants when Q&A is scripted. I thank Richard Frankel my dissertation committee chair for his guidance and mentorship. I also thank Gauri Bhat, Andrew Call, Ted Christensen, John Donovan, Bryan Graden, Jared Jennings, Chad Larson, Xiumin Martin, Lorien Stice-Lawrence, and Jake Thornock for their helpful comments. In addition, I thank workshop participants at Washington University in St. Louis and the Accounting Research Symposium at Brigham Young University.

2 1. Introduction Numerous books and articles describe techniques for predicting firm fundamentals using quantitative information found in firm disclosures. Recent academic studies find that qualitative disclosures also inform the market about future firm performance. For example, market participants gain useful information for predicting future performance by analyzing the tone (i.e., net optimistic language) of news stories (Tetlock et al. 2008), annual reports (Loughran and McDonald 2011), earnings press releases (Davis, Piger, and Sedor 2012), and earnings conference calls (Davis, Ge, Matsumoto, and Zhang 2012; Price et al. 2012). These studies argue that disclosure tone provides a signal of managers or others perceptions of firm fundamentals. This paper examines whether market participants use an alternative qualitative signal from quarterly earnings conference calls to predict firm fundamentals. Specifically, I test whether market participants infer negative information about future firm performance when firms script responses to questions received during the question and answer (Q&A) session of the earnings conference call. The Q&A session of the conference call is a unique setting in which managers and investors interact in two-way communication. While this interactive communication increases the flow of information to the market (Tasker 1998; Frankel et al. 1999; Bowen et al. 2002; Bushee et al. 2003; Matsumoto et al. 2011), research examines the potential disadvantages to the firm of allowing analysts and investors to ask questions in an open forum. For example, Hollander et al. (2010) suggest that the impromptu format of the Q&A session enables investors to prompt managers to reveal information they do not yet wish to reveal. 1 The potential for this type of inadvertent 1 Matsumoto et al. (2011) corroborate this argument. They find greater information content for the Q&A session of the conference call relative to both the presentation session and the accompanying press release, suggesting that some disclosures would perhaps not have been made were it not for questioning by analysts. 1

3 disclosure provides incentives for firms to prepare scripted responses to anticipated questions in advance of the conference call. I argue that firms are likely to script their Q&A when future performance is poor to avoid inadvertently providing information that can be used in litigation against the firm. Research suggests that when future performance is poor, litigation risk is high, and disclosure is costly. For example, Cutler et al. (2013) find evidence that firms with greater disclosure during the litigation class period are more likely to receive significant settlements against the firm. In addition, Rogers and Van Buskirk (2009) find that firms reduce disclosure following class action lawsuits, consistent with significant costs of disclosure in the litigation process. I similarly argue that firms holding regular earnings conference calls are likely to avoid the costs of providing information that can be used to build a case against the firm by preparing more careful and scripted responses to anticipated questions in advance of the conference call. If so, Q&A scripting can serve as a signal of negative future firm performance. However, if firms are more forthcoming with disclosure to prevent litigation as other research suggests (Skinner 1994; Kasznik and Lev 1995), scripting is unlikely to precede negative future events and would not provide a signal to market participants. Firms may also script responses to anticipated questions for reasons unrelated to future firm performance. For example, firms may script their responses to avoid inadvertently revealing proprietary information about the firm s products. Alternatively, managers who are less confident in their ability to respond to analysts questions in real time may use scripted responses to avoid the reputational costs of providing a botched answer to an analyst s question. Hence, market participants will interpret Q&A scripting negatively only if they assign a higher probability that firms script for expected performance reasons rather than to avoid proprietary or reputational costs. 2

4 I develop a measure of scripting based on linguistics research in computational stylistics that compares the stylistic properties of texts to determine authorship. I specifically examine the difference in the speaking style of the CEO during the presentation and Q&A sessions of the call and argue that CEOs who change their speaking style during the Q&A session are less likely to be relying on a script to respond to analysts questions. The implicit assumption in this measure is that the presentation session of the call is scripted and only partially prepared by the CEO. 2 Indeed, the investor relations team typically drafts the script and the CEO makes edits as necessary. Thus, the difference in the CEO s speaking style between the presentation session and the Q&A session identifies whether the CEO is using his/her unique style to answer questions or is relying on a script prepared by other individuals at the firm. Using a sample of 30,773 quarterly earnings conference call transcripts for 2,384 firms over the period from 2002 to 2011, I test the association between my measure of scripting and measures of future firm accounting performance and the market s response to the conference call. I find a negative association between Q&A scripting and both return on assets and operating cash flows in the four quarters subsequent to the conference call. These results are robust to using measures of unexpected future earnings and suggest that firms script their Q&A responses when they possess negative information about future firm performance. I also provide evidence that firms script their Q&A immediately prior to receiving a class action lawsuit consistent with firms scripting prior to bad news events. 2 Discussions with a former member of the internal investor relations team at Morgan Stanley verify that the presentation session of the call is scripted and is prepared by the investor relations team. Responses to expected questions are also often scripted. The member of the investor relations team estimated that the team is able to anticipate roughly eighty percent of the questions and draft prepared responses prior to the call. He also verified that the executives always read from the prepared script for the presentation session but often go off script during the Q&A session. However, for certain questions (such as a question received about level 3 fair value measurements) the executives respond from the prepared script word for word. 3

5 I next test the market reaction to scripted conference calls to identify (1) whether investors are able to discern the level of scripting done by management when answering questions and (2) whether investors view scripted responses as a negative signal of future firm performance. Controlling for the news in the earnings announcement, I find that firms that script their Q&A have significantly lower size and book-to-market adjusted returns on the day of the conference call. In addition, to more precisely control for news in the current period earnings surprise, I use TAQ data for a sub-sample of calls for which I have actual start times and find a negative association between scripting and abnormal returns following the call. This suggests that the negative association between scripting and the abnormal return on the day of the call is not solely due to the negative association between scripting and the current period earnings surprise, but that scripting also serves as a signal of future firm performance. In additional analysis, I find that sell-side equity analysts make downward revisions to their earnings forecasts in the 30 days following scripted conference calls, corroborating the market return tests and suggesting that analysts incorporate the negative implications of scripted Q&A into their forecasts. Finally, the negative market reaction to scripted Q&A is consistent with two explanations: 1) investors interpret scripted responses as a negative signal of future performance or 2) managers use scripted responses to provide additional information about the negative expected performance. I attempt to identify the most likely explanation in two ways. First, I directly test whether firms provide additional information about future earnings when conference calls are scripted. I find evidence that firms are less likely to issue earnings guidance on the day of the conference call when their Q&A is scripted suggesting that firms provide less, not more, information. Second, if firms provide additional information during scripted conference calls, analysts are likely to have a richer information set allowing them to make more accurate forecasts of future earnings. I find, 4

6 however, that analyst forecast revisions following scripted conference calls are less, not more, accurate. Hence, firms are unlikely to use scripted conference calls as a means of providing additional information to the market. Rather, the negative market reaction to scripted calls is consistent with scripted calls providing a signal of future firm performance. This paper contributes to the literature that examines the linguistic features of firm disclosures to extract information about the firm. 3 Prior research finds that disclosure tone and vocal cues in conference call speech are informative about firms future performance (Davis, Ge, Matsumoto, and Zhang 2012; Price et al. 2012; Mayew and Venkatachalam 2012). Other research finds that deceptive speech during conference calls predicts accounting misstatements (Larcker and Zakolyukina 2012). I add to this literature by examining an alternative conference call feature scripting of the Q&A session and find that my measure of scripting is correlated with future accounting performance, the market reaction at the time of the call, managers guidance decisions, and analyst forecast properties following the call. This paper also contributes to the literature that examines whether conference calls provide material information to conference call participants. Prior research finds significant trading activity at the time of the call (Frankel et al. 1999; Bushee et al. 2003; Bushee et al. 2004, Lansford et al. 2009), improvements in analyst forecast accuracy following the call (Bowen et al. 2002), more timely incorporation of earnings news into prices for firms initiating conference calls (Kimbrough 2005), and a reduction in information asymmetry for firms holding regular quarterly calls (Brown et al. 2004). This paper finds firms provide less information to market participants when conference calls are scripted. 3 Examples include Li (2008) and Lehavy et al. (2011) who examine the readability of financial reports, Brown and Tucker (2011) who examine firms year-over-year MD&A modifications, Li (2010) who examines forward looking statements in MD&A disclosures, and Tetlock et al. (2008), Loughran and McDonald (2011), Davis and Tama- Sweet (2012), Rogers et al. (2011), Davis, Piger, and Sedor (2012), and Blau et al. (2012) who examine disclosure tone in other settings. 5

7 The rest of the paper is organized as follows. Section 2 discusses the prior literature and develops the hypotheses. Section 3 outlines the empirical models used to test each hypothesis. Section 4 describes the sample selection process and summary statistics. Section 5 discusses the results. Section 6 examines whether firms provide more or less information during scripted calls. Section 7 provides sensitivity tests and additional analyses. Section 8 concludes the paper. 2. Background and Hypothesis Development Quarterly earnings conference calls have become an important form of voluntary disclosure. In 2002, approximately 17 percent of Compustat firms held at least one earnings conference call during the year and by 2011, the percentage increased to 36 percent (see Figure 1). For firms with analyst following, the percentages are much higher reaching 69% by In addition, firms that begin holding quarterly calls are likely to continue holding calls in the future. Hence, by implicitly committing to hold quarterly earnings conference calls, firms commit to a high level of transparency with the capital market. Conference calls generally involve two sessions: a presentation session in which management discusses results of operations for the quarter and a question and answer (Q&A) session in which analysts and investors ask questions of management. The conference call Q&A session is a unique voluntary disclosure setting in which managers and investors interact in twoway communication. Other forms of voluntary disclosure (e.g., press releases) are more one-sided. By allowing investors to ask questions, firms allow for the possibility that managers inadvertently reveal information the firm would have otherwise chosen to keep private. If disclosure of certain 4 These percentages include all firms on Compustat regardless of whether they have unique Factiva identifiers. If I restrict the focus to Compustat firms with Factiva identifiers, the percentages are much higher 77 percent in 2011 (85 percent for firms with analyst following). Thus, it is possible that these statistics understate the true number of firms holding conference calls since Factiva may not cover all firms on Compustat. 6

8 information is costly, firms can script their responses to anticipated questions as a means of providing more careful disclosure to outsiders. Firms typically employ an investor relations team to prepare a script for the presentation session with management providing edits and comments as necessary. The final script, therefore, is more likely to reflect the style of the investor relations team that prepared it, rather than the executive who eventually reads it over the call. The Q&A session, on the other hand, is more open and is considered a less scripted portion of the call (see, e.g., Matsumoto et al. 2011). However, if managers can anticipate or even prompt participants to submit questions prior to the call, investor relations teams can prepare scripted responses to these questions. Indeed, investor relations consultants often encourage firms to prepare for questions prior to conference calls. For example, Westwicke Partners, an investor relations firm, in a recent blog entitled Best Practices of Earnings Conference Call Preparation provide the following guidance: Compile the questions you expect to hear during the call Q&A Survey your sell-side analysts beforehand to learn what they are likely to ask. 5 I argue that firms are most likely to script responses to anticipated questions when future firm performance is poor. Firms with poor expected performance are subject to greater litigation risk and are likely more careful about the disclosures they make to external market participants since disclosures are often cited in class action lawsuits. For example, Cutler et al. (2013) find evidence that greater disclosure during the litigation class period results in a higher likelihood of significant settlements against the firm. Rogers and Van Buskirk (2009) also find evidence that firms reduce disclosure following class action lawsuits suggesting that disclosure is costly during litigation. Hence, firms are likely to use more careful and scripted disclosure when future 5 7

9 performance is poor to avoid the possibility of inadvertently revealing information that can be used to build a case against the firm. My first hypothesis is stated in the alternative form as follows: H1: Firms prepare scripted responses to anticipated questions for the conference call Q&A session when future firm performance is poor. My first hypothesis is less likely to hold if firms improve disclosure to prevent litigation as some research suggests. For example, Skinner (1994) and Kasznik and Lev (1995) find that firms are more likely to issue earnings guidance prior to periods of large negative earnings surprises relative to periods of large positive earnings surprises to avoid large negative market reactions at the earnings announcement date. In addition, Baginski et al. (2002) find that U.S. firms are more likely to issue earnings forecasts during periods of earnings declines relative to Canadian firms that operate in an environment where securities laws and judicial interpretations create a lower threat of litigation. In addition, my first hypothesis is less likely to hold if expected litigation costs are small or if firms believe scripting is unsuccessful in preventing significant settlements. Whether conference call Q&A scripting is negatively associated with future firm performance is, therefore, an empirical question. My second hypothesis examines investors response to scripted earnings conference calls as a joint test of (1) whether investors discern the level of scripting and (2) whether investors interpret scripted responses as a signal that managers possess negative information about future firm performance. Prior research suggests investors glean useful information from conference calls. For example, investors respond to managers conference call tone (Davis, Ge, Matsumoto, and Zhang 2012; Price et al. 2012) and to positive and negative affective states in vocal cues from conference call speech (Mayew and Venkatachalam 2012). Other research suggests investors respond negatively when managers refuse to answer specific questions during the Q&A session 8

10 (Hollander et al. 2010). If firms script their calls prior to periods of negative firm performance, and investors are able to discern whether managers are responding to questions from a script, I expect a negative market response to these calls. My second hypothesis is stated as follows: H2: Investors interpret scripted Q&A responses negatively. 3. Research Design 3.1 Conference call Q&A scripting measure The empirical challenge of this paper is identifying cross-sectional variation in the extent of conference call Q&A scripting. I develop my scripting measure using a computational stylistics method developed in the linguistics literature to identify the authors of documents with unknown or disputed authorship (see, e.g., Stramatatos 2009). The most well-known authorship attribution studies use linguistic methods to ascertain who wrote twelve of the Federalist Papers in which both Alexander Hamilton and James Madison claim authorship (Mosteller and Wallace 1963; Koppel, Schler, and Argamon 2009). Prior research suggests that the most effective method for authorship attribution is the comparison of a set of function words between two documents (Burrows, 1987; Stramatatos 2009; Mosteller 2010). Function words are those with primarily grammatical functions and include articles (e.g., a, an, the), conjunctions (e.g., and, or, so), pronouns (e.g., I, me, we), prepositions (e.g., of, on, in), and auxiliary verbs (e.g., is, do, can). 6 Mosteller (2010) suggests that function words are the best stylistic discriminators between two authors because they are unrelated to the topic discussed, and they reflect minor or even unconscious preferences of the author. Thus, an author s use of function words uniquely identifies 6 See Appendix A for a complete list of function words used in this study. 9

11 his/her style. Using this approach, studies overwhelmingly identify James Madison as the author of the twelve disputed Federalist Papers (Mosteller and Wallace 1963). 7 Using this method, I examine the extent of scripting of the Q&A session of the conference call by comparing the use of function words by the CEO during the presentation session to the use of function words by the CEO during the Q&A session. 8 I assume the presentation session of the call is a scripted outline of the performance of the firm during the quarter. Conversations with an investor relations consultant and a member of the internal investor relations team at Morgan Stanley confirm this assumption. The set of function words during this session of the call thus serves as a baseline for which I can compare the set of function words during the Q&A session of the call. A CEO is less likely to be relying on scripted responses to conference call questions if the use of function words during the Q&A session is less similar to the use of function words during the presentation session of the call. In other words, if the CEO s speaking style changes from the presentation session to the Q&A session, he/she is less likely to be using a script to respond to analysts and investors questions. For each conference call, I first identify the presentation and Q&A sessions of the call by searching for key words such as question and Q&A within 2 lines of other key words such as take or open up. 9 I then identify the chief executive officer using the titles provided during the call and obtain the portions of the call in which the executive is speaking. 10 Next, I create two 7 Other methods used in prior work include comparing sentence lengths, word lengths, or uses of frequent words between two documents. However, these methods are shown to be poor indicators of authorship (see Mosteller 2010). For this reason, I use the most accepted approach of comparing function words between two documents. 8 The results of all tests remained qualitatively and quantitatively similar if I use the spokesman executive to compute the scripting measure where the spokesman is defined as the CEO or CFO who speaks for the longest portion of the conference call. See Section 6.1 for additional detail. 9 During the introduction of the call, the executives often provide an outline for the call and state they will be opening up the call for questions later on in the call. To ensure I obtain the key words when the Q&A session truly begins rather than a reference to it later in the call, I require the Q&A session to start at least 10% into the call. 10 In many instances, the conference call speaker is identified using an abbreviated version of the executive s name. For example, the executive might be referred to as David when introduced but then Dave later in the call. I manually correct these differences to ensure I obtain the full text of the call for each executive. 10

12 vectors of the counts of the function words spoken by the CEO in each session of the call: v QA and v PRES, respectively, where QA represents the Q&A session and PRES represents the presentation session. I then compute my measure of scripting as the cosine similarity between the two vectors using the following formula: SCRIPT = cos(θ) = v QA v PRES v QA v PRES (1) where θ is the angle between v QA and v PRES, ( ) is the dot product operator, and v i is the length of vector v i (i is equal to QA and PRES). The cosine similarity measure captures the uncentered correlation between two vectors and provides an estimate of the similarity in the use of function words by the executive during the presentation and Q&A sessions of the conference call. 11 Its values range between 0 and 1 where greater values indicate greater similarity. For ease in economic interpretation in the multivariate analyses, I rank the SCRIPT measure into deciles from 0 to 9 and divide by 9 (RSCRIPT). 12 I also require at least 200 words to be spoken by the CEO in both the presentation session and the Q&A session of the call to reduce measurement error. I verify the construct validity of the cosine similarity measure in identifying the speaking style of the CEO by computing the cosine similarity measure between the vector of function word counts spoken by CEO j during the Q&A (presentation) session for firm i in quarter t to the vector of the combined conference call Q&A (presentation) sessions given by CEO j for firm i during all other quarters. I then compute the cosine similarity between the CEO j Q&A (presentation) function word count vector in quarter t to nine randomly selected combined word count vectors for CEOs of other firms across the sample period. I then rank the actual CEO vector relative to 11 Brown and Tucker (2011) use the cosine similarity measure to compare firms MD&A disclosures over time. Their word count vectors include all unique words in the disclosure to compare content, whereas I use only the counts of function words to compare speaking style. 12 The results remain qualitatively unchanged if I use the unranked cosine similarity measure. 11

13 the nine randomly selected CEO vectors, where values of 1 (10) indicate the actual CEO vector is the most (least) similar relative to the nine randomly-selected CEO vectors. Figure 2 presents the cumulative percentage of firms in each ranking. If the ranking were random, the percentage of firms in each ranking would be 10 percent. When comparing the Q&A session during the quarter to the Q&A sessions of other quarters (Q&A to Q&A), the results indicate that 79.7 percent of the similarity scores are highest for the actual CEO relative to the nine randomly selected CEOs. The similarity score for the actual CEO is one of the top three highest for 93.3 percent of the observations suggesting that the similarity score does a good job of identifying the speaking style of the CEO. Similarly, when comparing the presentation session during the quarter to the presentation sessions of other quarters (PRES to PRES), the results indicate that 80.5 percent of the similarity scores are highest for the actual CEO relative to the nine randomly selected CEOs suggesting that those who script the presentation session (e.g., the investor relations team) have uniquely identifiable styles. 13 I then compute the cosine similarity between the presentation session vector for CEO j of firm i in quarter t and 1) the Q&A session vector for CEO j of firm i in quarter t and 2) nine randomly-selected Q&A session vectors for CEOs of other firms. I then rank the similarity score for the actual CEO vector relative to the randomly-selected CEO vectors. Figure 2 plots the 13 I further verify the accuracy of the cosine similarity measure in the most common setting used in the linguistics literature: The Federalist papers. I compute the cosine similarity between the vector of word counts for each Federalist paper and the vectors of word counts for the three known authors of the Federalist papers: John Jay, James Madison, and Alexander Hamilton. I assign an author to each paper based on the highest similarity score for each paper relative to the vectors of word counts for all other papers written by the three authors. For all five papers written by John Jay, the similarity score correctly identifies John Jay as the author. For the 51 papers known to have been written by Alexander Hamilton, the similarity score correctly identifies 48 as written by Hamilton and incorrectly identifies 3 as written by Madison. For the 14 papers known to have been written by James Madison, the similarity score correctly identifies 12 as written by Madison and incorrectly identifies 2 as written by Hamilton. For the 12 disputed papers, I find 10 of the similarity scores are highest for James Madison and 2 of the similarity scores are highest for Alexander Hamilton. These results are fairly consistent with prior research and provide additional evidence that the similarity score using the list of function words employed in this study provides an accurate measure for detecting subtle differences in style between two texts. 12

14 cumulative percentage of conference calls in each ranking (PRES to Q&A). I find that only 21.4 percent of the similarity scores are highest for the Q&A session of the actual CEO compared to the nine randomly-selected Q&A sessions of other CEOs. This suggests two important points. First, CEOs have unique styles relative to the investor relations teams that prepare the presentation sessions of the calls. If not, the percentage of firms with rankings closer to 1 would have been closer to 100 percent. Second, the percentage of firms with a ranking of 1 is greater than what would be expected if the rankings were random (21 percent relative to 10 percent) suggesting that some firms script their Q&A. 3.2 Test of hypothesis one I test the association between Q&A scripting and firms future accounting performance (Hypothesis 1) by estimating the following model similar to Core, et al. (1999), Bowen et al. (2008), and Davis, Piger, and Sedor (2012): FUT PERFi,t = α0 + α1 RSCRIPTi,t + α2 PERFi,t + α3 EARN SURPi,t + α4 ln(mvei,t) + α5 INSTOWNi,t + α6 ln(anal FOLLi,t) + α7 TURNOVERi,t + α8 EARN VOLi,t + α9 RET VOLi,t + α10 ln(agei,t) + α11 GUIDANCEi,t + α12 GUID SURPi,t + α13 TONEi,t + α14 ln(ceo WC PRESi,t) + α15 ln(ceo WC QAi,t) + YEARQTR + INDUSTRY + εi,t. (2) The dependent variable, FUT PERFi,t, is the average accounting performance of firm i over the four quarters following quarter t. I examine two measures of future accounting performance: FUT ROAi,t and FUT CFOi,t, where FUT ROAi,t (FUT CFOi,t) is the average income before extraordinary items (operating cash flow) divided by lagged total assets for firm i over the four quarters following quarter t. The independent variable of interest is the RSCRIPTi,t variable which is the conference call Q&A scripting measure defined in section 3.1. I expect a negative 13

15 association between FUT PERFi,t and RSCRIPTi,t if firms script Q&A responses when they possess negative information about future firm performance. I include several additional firm-specific variables to control for factors likely associated with Q&A scripting and future firm performance. I first include the current value of PERFi,t to control for persistence in the performance measures (Barber and Lyon 1996). PERFi,t is measured as return on assets (ROAi,t) or cash flow from operations scaled by lagged total assets (CFOi,t) for firm i in quarter t when FUT ROAi,t and FUT CFOi,t are the dependent variables, respectively. Next, I include the natural logarithm of market value of equity (ln(mvei,t)) to control for firm size and expect larger firms have higher future accounting performance (Core et al. 1999). I also include the earnings surprise for firm i in quarter t (EARN SURPi,t) defined as IBES actual EPS less analysts median consensus forecast prior to the conference call date divided by share price at the end of the quarter and expect a negative coefficient consistent with Davis, Piger and Sedor (2012). I also include the standard deviation of earnings in the previous 16 quarters (EARN VOLi,t) and the standard deviation of monthly stock returns over the previous twelve months (RET VOLi,t) to control for firm risk. Consistent with prior research, I expect a negative association between future performance and risk (Minton et al. 2002; Bowen et al. 2008; Core et al. 1999; Davis, Piger, and Sedor 2012). I also control for the firm s life cycle stage by including firm age (ln(agei,t)) defined as the natural logarithm of the number of years since the firm first appeared on Compustat as of quarter t. I expect younger firms have lower future performance. I control for additional factors potentially associated with my scripting measure: the percentage of institutional ownership of the firm (INST OWNi,t), the natural logarithm of the number of analysts following the firm during the quarter (ln(anal FOLLi,t)), and the stock turnover for the firm during the quarter 14

16 (TURNOVERi,t) defined as trading volume divided by the number of shares outstanding. I also include an indicator variable equal to 1 if the firm provides earnings guidance for the next quarter s EPS on the conference call date and 0 otherwise (GUIDANCEi,t) to control for manager s provision of quantitative information during the call. I use guidance data from both First Call and IBES to reduce issues associated with the completeness of the datasets (Chuk et al. 2012). I also control for the direction of the surprise in the earnings guidance according to First Call and IBES by defining a variable equal to 1 if the guidance qualifies as a positive earnings surprise, equal to 0 if the guidance does not qualify as a surprise, and equal to -1 if the guidance qualifies as a negative surprise (GUID SURPi,t). When the firm does not provide earnings guidance, GUID SURPi,t is set equal to 0. I next include several conference call specific variables. I first include conference call tone (TONEi,t) defined as the number of positive words less the number of negative words in the call divided by the total number of words in the call. The positive and negative word dictionaries are obtained from Loughran and McDonald (2011). I expect a positive association between tone and future performance if net optimistic language indicates positive information about future performance. I next include the natural logarithm of the number of words spoken by the CEO during the presentation session (ln(ceo WC PRESi,t)) and during the Q&A session of the call (ln(ceo WC QAi,t)) to control for potential measurement error in the scripting measure if larger word counts provide a more precise measurement of the differences in function words between the presentation and Q&A sessions of the conference call. Finally, I include year-quarter and industry (two-digit SIC code) indicator variables to control for differences in Q&A scripting over time and across industries. I also cluster the standard 15

17 errors by firm due to likely serial correlation in the dependent and independent variables (Petersen, 2009). 3.3 Test of hypothesis two To test whether scripted conference calls are associated with a negative stock market reaction at the conference call date (Hypothesis 2), I estimate the following model similar to Mayew and Venkatachalam (2012): CC CARi,t = δ0 + δ1 RSCRIPTi,t + δ2 EARN SURPi,t + δ3 ROAi,t + δ4 ln(mvei,t) + δ5 BTMi,t + δ6 MOMi,t + δ7 RET VOLi,t + δ8 GUIDANCEi,t + δ9 GUID SURPi,t + δ10 TONEi,t + δ11 ln(ceo WC PRESi,t) + δ12 ln(ceo WC QAi,t) + YEARQTR + INDUSTRY + εi,t. (3) The dependent variable is the size and book-to-market adjusted cumulative abnormal stock return over the window [0,1] surrounding the earnings conference call date (CC CARi,t). The stock is matched to one of the 25 size-btm Fama French portfolios based on the market capitalization of the firm at the end of June and the book value of equity of the last fiscal year end in the prior calendar year divided by the market value of equity at the end of December of the prior year. 14 The independent variable of interest is the RSCRIPTi,t measure defined in section 3.1. I expect a negative association between RSCRIPTi,t and CC CARi,t if investors interpret Q&A scripting as a negative signal of future firm performance (Hypothesis 2). I control for size, growth, and risk, which have been shown to be related to market returns (Collins and Kothari 1989). I use the natural logarithm of market value of equity (ln(mvei,t) as the proxy for size, book-to-market (BTMi,t) as the proxy for growth, and return volatility (RET VOLi,t) as the proxy for risk. In addition, I control for return momentum (MOMi,t) defined as the 14 The breakpoints and 25 portfolio returns are obtained from Kenneth French s website at 16

18 cumulative abnormal return over the [-127, -2] window prior to the conference call date. I also include the current period earnings surprise (EARN SURPi,t) and the current period return on assets (ROAi,t) to control for the market reaction to current period earnings and expect a positive coefficient on these variables. I also include the earnings guidance variables GUIDANCEi,t and GUID SURPi,t to control for quantitative information provided by the firm about future performance. I also include the conference call specific variables TONEi,t, ln(ceo WC PRESi,t), and ln(ceo WC QAi,t) to control for alternative linguistic features of the conference call and for potential measurement error in the scripting measure. Consistent with prior research, I expect a positive association between conference call tone and the market reaction to the call. Finally, I include year-quarter and industry (two-digit SIC code) indicator variables and cluster the standard errors by firm. 4. Sample selection and data I obtain a sample of earnings conference calls by first matching all non-financial firms on Compustat with non-missing total assets between 2002 and 2011 to their corresponding unique Factiva identifiers using the company name provided by Compustat. 15 For the 11,702 unique Compustat firms, I find Factiva identifiers for 5,099 firms. Using each firm s unique identifier, I then search Factiva s FD Wire for earnings conference calls made between 2002 to 2011 and find 56,822 total calls for 3,475 unique firms. 16 I remove 15,384 calls in which the CEO speaks less than 200 words in either the presentation or Q&A session of the call. Requiring financial statement data from Compustat, IBES, and CRSP further reduces the sample by 5,142 calls, 1, In cases where the match is ambiguous, I check whether the city and state of the matched firm in Factiva matches the city and state of the firm in Compustat. 16 Factiva contains different types of conference calls such as those discussing mergers and acquisitions. I focus only on earnings-related conference calls. I filter out non-earnings related conference calls by requiring the term earnings to be in the title of the call. I also require the conference call be made within 2 days of the earnings announcement. 17

19 calls, and 3,813 calls, respectively. The final sample consists of 30,773 earnings conference calls for 2,384 unique firms with sufficient data to estimate the main empirical analyses. Table 1 presents the means of the variables used in the empirical analysis for the full sample and also for each quintile of the SCRIPTi,t measure. The final column in the table reports the test statistic testing the difference between the fifth and first quintile. The mean of the scripting measure (SCRIPTi,t) is in the bottom quintile and in the top quintile. The mean of future return on assets (FUT ROAi,t) is in the bottom scripting quintile and in the top quintile and the mean of future operating cash flows (FUT CFOi,t) is in the bottom quintile and in the top quintile and the differences are statistically significant at the one percent level providing preliminary evidence of a negative association between Q&A scripting and future performance (Hypothesis 1). Table 1 also reports a significant difference in the cumulative abnormal return at the conference call date (CC CARi,t) between the top and bottom quintiles of the scripting measure ( compared to 0.005) providing preliminary evidence that investors interpret scripting as a signal that mangers possess negative information about future firm performance (Hypothesis 2). The cumulative abnormal return in the 252 trading days following the conference call (FUT CARi,t) shows no difference between the top and bottom quintiles, suggesting that investors understand the implications of Q&A scripting and there is no drift. I also find analyst forecast revisions following the conference call (FREVi,t+1) are more negative in the top quintile of the scripting measure relative to the bottom quintile ( compared to ). I also find that 19.9% of firms in the top quintile of the scripting measure provide guidance for next quarter s EPS (GUIDANCEi,t) compared to 22.8% in the bottom. I do not, however, find a difference in analyst forecast accuracy (ACCURACYi,t+1) between the top and bottom quintiles. 18

20 I also report the means of the control variables used in the empirical analysis. The significant differences between the top and bottom quintiles for these variables underscore the importance of including these variables in the empirical analysis to control for alternative explanations. I specifically find that firms in the top quintile are larger with greater analyst following and institutional ownership, have lower current period market and accounting performance, have lower book-to-market ratios, have been listed on Compustat for less time, and provide more negative forecasts of future EPS. I also find that firms with CEOs who speak more words during both the presentation and Q&A sessions have more scripted conference calls which can be attributable to two forces. First, when the firm does not wish to inadvertently disclose information to outsiders, it may script longer presentations and responses to analysts questions to allow for less time for multiple questions to be asked. Second, higher word counts allow a more precise measurement of the scripting variable potentially creating a bias in the measure. Hence, I include these two measures in each regression specification to control for this possibility Results 5.1. Results for hypothesis one Table 2 presents the results of estimating Equation 2. In Column 1 (2) the dependent variable is FUT ROAi,t (FUT CFOi,t). The coefficient on RSCRIPTi,t is in Column (1) and in Column (2) and both are significant at the one percent level. The coefficient estimates suggest that relative to firms in the bottom decile of the scripting measure, firms in the top decile have a 45 percent lower return on assets in the four quarters following the conference call (- 17 To further rule out the possibility that measurement error in the scripting measure is affecting my results, I reestimate the scripting measure holding the number of words constant across firms. I continue to find a highly positive correlation between this alternative scripting measure and the total number of words spoken by the CEO during both the presentation and Q&A sessions of the call, suggesting measurement error is not driving the large positive association between call length and my scripting measure. The results of my empirical analyses are also robust to using this alternative measure. See Section 6.1 for more detail. 19

21 0.003/ = -0.45) and 21 percent lower operating cash flows in the four quarters following the conference call (-0.003/ = -0.21). These results suggest that firms script Q&A when future accounting performance is poor and are consistent with my first hypothesis. The control variables indicate that larger firms with more institutional ownership, lower return volatility, and lower analyst following have higher future earnings and cash flows. I find positive coefficients on the current period performance measures consistent with persistence in performance. I also find that younger firms with higher stock turnover and lower earnings volatility have higher future cash flows but that these variables are insignificant in the future earnings regression. In addition, future earnings and cash flows are higher when firms provide guidance and when the guidance is more positive. I also find that conference call tone loads positively in both future performance regressions, suggesting that managers use positive tone when future performance is high. I do not find a relation between future performance and the number of words spoken by the CEO during the Q&A session, but lower future earnings when the presentation session is longer Results for hypothesis two I next estimate the relation between scripting and the market reaction at the time of the conference call. Panel A of Table 3 presents the results of estimating Equation (3). The coefficient on RSCRIPTi,t is and significant at the one percent level in Column (1) without including the control variables. After including the control variables in Column (2), the magnitude of the coefficient drops to but remains statistically significant at the one percent level. The coefficient in Column (2) indicates that relative to firms in the bottom decile, firms in the top decile of RSCRIPTi,t have 139 percent lower abnormal returns at the conference call date relative to the mean of CC CARi,t (-0.003/ = -1.39). This result is consistent with investors interpreting 20

22 scripted calls as a negative signal of future performance and supports my second hypothesis. The control variables indicate that larger and higher growth firms have lower conference call returns. I also find a negative relation between the conference call return and return momentum. Firms with more positive ROA and more positive earnings surprises also have higher abnormal returns. I also find that firms with more positive earnings guidance on the day of the call have higher abnormal returns on the day of the call, but that the decision to guide future earnings is negatively associated with the abnormal return. In addition, firms with more positive conference call tone have higher abnormal returns consistent with prior research. I also find that firms with longer presentation sessions have lower abnormal returns. I next examine whether scripted conference calls are associated with future abnormal returns to determine whether investors over or under react to scripted calls at the conference call date. Panel B reports the results of Equation (3) replacing CC CARi,t with FUT CARi,t, defined as the abnormal return for the 252 trading days following the conference call using the window [2, 254]. In Column (1) I do not find a significant relation between RSCRIPTi,t and FUT CARi,t, suggesting that the reaction at the conference call date does not reverse in future periods on average, and hence, was not an overreaction. Instead, scripted calls provide investors with a signal of future negative performance at the conference call date. However, some firms are likely to script their Q&A responses for reasons unrelated to future performance. For example, proprietary costs of inadvertent disclosure can induce some firms to script their Q&A responses to avoid revealing information about the firm s products. In addition, managers who are less confident in their ability to respond to questions are likely to rely on scripted responses to avoid tarnishing their reputational capital. For these firms, when future performance materializes and the market s negative prior assessment of performance is proved 21

23 inaccurate, the negative stock market response is likely to reverse. I test this conjecture by including the interaction between RSCRIPTi,t and an indicator variable equal to 1 for below median values of FUT ROAi,t (LOW FUT ROAi,t) as an additional control variable in the second column of Panel B. I expect a positive coefficient on RSCRIPTi,t if returns reverse for firms with high subsequent performance. In contrast, I expect the sum of the coefficients on the RSCRIPTi,t measure and the interaction between RSCRIPTi,t and LOW FUT ROAi,t to be insignificant if returns do not reverse for firms with poor subsequent performance. The results in Column (2) are consistent with these expectations. The coefficient on RSCRIPTi,t is and is significant at the one percent level. In contrast, the sum of the coefficients on the RSCRIPTi,t measure and the interaction between RSCRIPTi,t and LOW FUT ROAi,t is and is insignificant. Next, I corroborate the results in Panel A of Table 3 by examining revisions of analysts EPS forecasts for quarter t+1 following the conference call date. Specifically, I regress analyst forecast revisions, FREVi,t+1, defined as the median analyst EPS forecast for quarter t+1 for all forecasts made within 30 days following the conference call date less the median consensus forecast of quarter t+1 directly prior to the conference call divided by price and multiplied by 100 on the scripting measure and other control variables. 18 Table 4 presents the results and reports a negative coefficient on RSCRIPTi,t of -0.10, which is significant at the one percent level. The coefficient estimate suggests that moving from the bottom to the top decile of the RSCRIPTi,t measure is associated with a 51 percent decrease in FREVi,t+1 relative to the mean of FREVi,t+1 (- 0.10/ = 0.51). This result is consistent with the abnormal returns tests and suggests that analysts revise downward their forecasts of future earnings after scripted conference calls. Thus, sophisticated investors (i.e., analysts) view conference calls as a negative signal of future firm 18 I multiply by 100 to be able to observe the coefficient on the scripting variable without reporting several decimal places. 22