Negative News and Investor Trust: The Role of $Firm and #CEO Twitter Use

• Date: 01 December 2018
• Published date: 01 December 2018
• Author: FRANK D. HODGE,W. BROOKE ELLIOTT,STEPHANIE M. GRANT
• DOI: http://doi.org/10.1111/1475-679X.12217

DOI: 10.1111/1475-679X.12217

Journal of Accounting Research

Vol. 56 No. 5 December 2018

Printed in U.S.A.

Negative News and Investor Trust:

TheRoleof$Firmand#CEO

Twitter Use

W. BROOKE ELLIOTT,

∗STEPHANIE M. GRANT,

†

AND FRANK D. HODGE

†

Received 1 September 2016; accepted 15 April 2018

ABSTRACT

We examine how CEOs can facilitate the development of investor trust that

helps mitigate the effects of negative information. Results from an experi-

ment show that investors trust the CEO more and are more willing to invest in

the firm when the CEO communicates firm news followed by a negative earn-

ings surprise through a personal Twitter account than when the news and sur-

prise comes from the CEO via a website or from the firm’s Investor Relations

Twitter account or website. A follow-up experiment shows that repeating the

negative news does not incrementally affect investors who received the news

from the CEO’s Twitteraccount, but does further negatively impact investors

∗University of Illinois at Urbana–Champaign, Gies College of Business; †University of

Washington, Foster School of Business.

Accepted by Douglas Skinner. The authors would like to thank an anonymous re-

viewer, Spencer Anderson, Jason Brown, Tim Brown, Joe Burke, Clara Chen, Deni Cikurel,

Kirsten Fanning, Joe Fisher, Brian Gale, Brent Garza, Shawn Gordon, Kamber Hetrick,

Sean Hillison, Jessen Hobson, Ryan Hudgins, Kim Ikuta, Elena Klevsky, Ethan LaMothe,

Justin Leiby, Jeremy Lill, Tracie Majors, Jamie Pratt, Kathy Rupar, Rosh Sinha, Laura Wang,

Shuyang Wang, Brian White, Amanda Winn, Ha Young Yoon, attendees at the 2015 AAA

Annual Meeting, attendees at the 2015 ABO conference, doctoral students at the Uni-

versity of Washington, and workshop participants at the University of Illinois at Urbana–

Champaign, Indiana University, and the University of Florida for helpful comments, and

Xavier Lange for programming assistance. Frank Hodge acknowledges support of the Michael

G. Foster Endowment. An online appendix to this paper can be downloaded at http://

research.chicagobooth.edu/arc/journal-of-accounting-research/online-supplements.

1483

CUniversity of Chicago on behalf of the Accounting Research Center,2018

1484 W.B.ELLIOTT,S.M.GRANT,AND F.D.HODGE

who received the news via other disclosure mediums, especially those who

received the news via the Investor Relations Twitter account. Our results have

implications for firms and executives considering the costs and benefits of

communicating with investors via Twitter.

JEL codes: G11; G40; M41

Keywords: social media; Twitter; social bond; trust; investment decisions;

negative earnings surprise; repeated information

1. Introduction

Firms increasingly use social media to disclose financial information (Joyce

[2013], Blankespoor, Miller, and White [2014]), and investors increasingly

use social media to retrieve firm information (Alden [2013], Brunswick

[2014]). One of the most popular social media disclosure platforms is Twit-

ter. Using Twitter to disclose financial information introduces two unique

features not present in traditional disclosure platforms. First, Twitter cre-

ates an opportunity for investors to develop a social bond with whomever is

tweeting out the information. Second, Twitter increases the repetition of in-

formation through multiple tweets and retweets of selected content (Miller

and Skinner [2015]). We examine how these features have the potential to

facilitate the development of investor trust that helps mitigate the effects of

negative information.

The specific information we focus on is negative earnings surprises. We

do so because prior research suggests these events can have significant

consequences, for example, a disproportionately large stock price decline

(Skinner and Sloan [2002]), loss of investor trust (Graham, Harvey, and

Rajgopal [2005]), and negative career outcomes for the managers (Mer-

genthaler, Rajgopal, and Srinivasan [2012]). As a result, managing

negative earnings surprises is important to managers (Graham, Harvey,

and Rajgopal [2005]). Our study is important because it sheds light on

how managers can mitigate investors’ loss of trust and negative reaction

that follows a negative shock. In the first of two experiments, we examine

if investors develop more enduring trust in a firm’s CEO when that CEO

communicates with them via Twitter. We further examine whether this trust

helps temper investors’ reaction to subsequent negative firm news. In the

second experiment, we examine if the trust investors developed is robust

to the repetition of negative news. In addition, our second experiment has

the incremental benefit of providing a potential explanation for the finding

that the market is more sensitive to negative earnings surprises when firms

are active on Twitter (e.g., Curtis, Richardson, and Schmardebeck [2016],

Jung et al. [2017]).

To examine our research questions, we draw on social identity theory

to predict that investors develop social bonds when another individual dis-

closes information via Twitter. The social bond is created by the perception

that investors are personally interacting with the other individual. A social

NEGATIVE NEWS AND INVESTOR TRUST 1485

bond refers to an individual’s subjective sense of interpersonal closeness or

connectedness with another (Kadous, Leiby, and Peecher [2013]). Social

bonds generally cause individuals to develop more enduring trust in the

other individual (Shapiro, Sheppard, and Cheraskin [1992], Lewicki and

Bunker [1996], Haslam and Ellemers [2005]). Importantly, the extent to

which the trust remains following negative news depends on the strength

of the social bond developed; the stronger the initial bond the more likely

trust remains intact after negative news.

We focus on trust at an operational level because (1) theory suggests so-

cial bonds are reflected in measures of trust and (2) trust has been shown

to matter in an investment context (see, e.g., Elliott, Hodge, and Sedor

[2012]). Since we focus on trust and the formation of a social bond, we

measure the two foundations of trust—affective and cognitive trust (Lewis

and Weigert [1985]). Mercer [2005, p. 724] examines a related, but dis-

tinct, construct of management credibility, which she defines as “investors’

beliefs about management’s trustworthiness and competence in financial

disclosure.” Credibility captures the cognitive foundation of trust. We mea-

sure that foundation as well as the affective foundation of trust, which is

theoretically important for the formation of social bonds.

We expect that the source of a message, and the medium used to com-

municate the message, interact to affect the strength of the social bond

that forms, which in turn affects investors’ trust. Specifically, we expect

that investors will develop a stronger social bond (and more enduring

trust) when a CEO communicates firm news via a personal Twitter ac-

count than when the firm’s Investor Relations communicates news via a

Twitter account or when news is communicated via a website. As a result

of having formed enduring trust, investor reactions to subsequent nega-

tive news will be tempered relative to when the same news is received via

other disclosure mediums. We follow up our primary analysis by examin-

ing the mediating role trust plays in influencing investor decisions in our

setting.

We test our predictions by conducting two experiments. In our primary

experiment, we isolate one of the two key features of Twitter: its ability to

facilitate a social bond. Doing so increases the internal validity of our exper-

iment, but at the same time, does not mimic how Twitter functions in the

real world. We therefore conduct a follow-up experiment that more closely

mimics Twitter in the real world, thereby increasing the generalizability of

our findings. Both experiments use 2 ×2 between-subjects designs with the

same independent variables: disclosure medium (website vs. Twitter) and

message source (Investor Relations vs. CEO).1

1Our study examines a potential benefit of CEOs using personal Twitter accounts to com-

municate firm news, but this choice could also introduce costs that we do not examine in our

study.As discussed in more detail in section 2, several legal and regulator y risks exist related to

social media accounts, including risks of violating advertising, intellectual property, and SEC

laws and regulations (e.g., Regulation FD).