Buying on rumors: how

financial news flows affect

the share price of Tesla

Nadine Strauss

Institut fur Publizistik und Kommunikationswissenschaft,

Universitat Wien, Vienna, Austria, and

Christopher Holmes Smith

Annenberg School for Communication and Journalism,

University of Southern California, Los Angeles, California, USA

information environment.

Design/methodology/approach – Using a multi-method case study approach, combining quantitative

intraday event studies with a qualitative text analysis of financial online news and tweets by Elon Musk

and Twitter, the authors shed light on the complex interaction between market events, financial information

and stock market reactions. The analysis covers a period of four days, encompassing the announcement and

introduction of the new battery pack for Model S and X by Tesla as well as the accompanying and follow-up

reporting by the financial news media.

Findings – Findings show that market reactions are driven by business events and expectations among the

market rather than the follow-up reporting by financial news media. Financial online news instead seems to

heavily rely on Elon Musk’s attention-triggering news to sustain its 24-h airtime with a variety of reporting

regarding insider-trading and the publication of market-relevant information. In addition, corporate financial

communication should avoid impetuous communication via social media channels as this could have

deterrent effects on the market valuation of a listed company.

Keywords Financial markets, Event study, Financial reporting, Stock markets, Market reactions,

Financial news

Paper type Research paper

The flow of information can severely impact the formation of stock prices (Mitra and Mitra,

2011). Similarly, media representations of companies might not only have a considerable

influence on corporate reputation (cf. Fombrun and Shanley, 1990), but might also affect

the fluctuations of their share prices (e.g. Scheufele et al., 2011). Scholars have argued that

Accepted 4 June 2019

about the company’s future business prospects, and hence its market valuation (Pollock and

Rindova, 2003). Previous research, however, has investigated the link between media

representations and the stock market from a rather static perspective, measuring, for

example, the amount of coverage and sentiment expressed in news items over a certain

amount of time and the stock market reactions therewith (e.g. Scheufele et al., 2011; Strauß

et al., 2016; Tetlock, 2007). Subsequently, studies in this research area have not paid

attention to the complex dynamics of event-driven news coverage and its effects on stock

market prices. It remains unclear how breaking issues can move the market gradually and

how various outlets report on the issues over time, affecting investors’ trading-decisions.

In addition, such event-driven news coverage increasingly originates in the realm of

media outlets, the timing of coverage, and therewith, the dynamic effects of changing

reporting on corporate news and stock market prices over time. Thus, the overarching

research question of this study reads:

RQ1. How do financial news flows about corporate events influence the stock market

prices of companies over time?

However, previous research has led to the “logical conclusion that there is no conclusive and

systematic relationship between big news and big price movements” (Schuster, 2006, p. 50).

Therefore, instead of looking at aggregate effects of corporate news on stock market prices

(e.g. Scheufele et al., 2011; Strauß et al., 2016; Tetlock, 2007), researchers are advised to take a

micro-level perspective when investigating how news flows affect stock market prices. In

Knorr Cetina and Bruegger (2002), who studied foreign-exchange dealing rooms through an

economic sociological lens, argue that “market reality itself is knowledge generated, that is,

has no existence independent from the informational presentation of the market on screen that

is provided by news agencies, analysts, and traders themselves” (p. 915). These information

flows exist within a crowded and turbulent global media ecosystem where the competition to

capture attention across linear broadcast feeds, digital on-demand services and social media

networks is fierce. In this “attention economy,” media users – including variously situated

social agents actively engaged in the screen-driven activities that constitute modern financial

spectacles” (Kellner, 2003) – that frequently erupt within the news cycle and shatter the

quotidian pattern of media reception, thereby eliciting more intense forms of coverage rife

with “endless detail and speculation about further details” on the unfolding story (p. 85).

When describing financial news flows in the global, attention-seeking and ubiquitous media

issues in the USA. The five stages are divided in: pre-problem state, alarmed discovery and

euphoric enthusiasm, realizing the costs of significant progress, gradual decline of public

explaining how public attention in America has slowly increased toward environmental issues,

Iyengar and Kinder, 1987; Shoemaker and Reese, 1996). Agenda-setting presumes that the

attention media pays toward certain issues gets transferred to the public agenda (McCombs and

Shaw, 1972). In other words, the more the media talks about a certain topic (e.g. Tesla product

introduction), the more likely the audience will recall this topic and rate it as important. Similarly,

framing theory suggests that public perceptions of characters (e.g. politicians), topics or

institutions are dependent on how the news media frame these (Carroll and McCombs, 2003;

Golan and Wanta, 2001). As a consequence, the framing of a certain topic, institution or person in

the media in a positive or negative light might impact the way the public perceives and

Hence, following agenda-setting and framing theory, financial news might also impact

emotions (e.g. Bollen et al., 2011) or expert opinions (e.g. Bar-Haim et al., 2011). While some

findings point to correlations between news, the sentiment of the coverage and the stock

market on an aggregate level (but are not generalizable), it is still open for discussion how

this relationship can be described at the micro-level; hence, when studying news dynamics

for a specific corporate event or a specific listed company.

Indeed, scholars in news analytics have highlighted the importance of news flows and the

fact that news can switch from positive to negative sentiment for a specific company over time

colleagues argue that a focus on short-term influences of news events on share prices is more

appropriate. Similarly, it is argued in this study that the accumulated and mostly one-way

Figure 1.

Distribution of

announcement

where AR

is the abnormal return for stock i (Tesla) at time point t (trading minute on

August 23, 2016); R

the actual return; and E(R

return of a comparable market portfolio X at time point t (MacKinlay, 1997). Given that for

this study the market model was chosen (MacKinlay, 1997), X is the market return of the

also the common estimation window employed in event studies (cf. MacKinlay, 1997). By

means of the parameter estimates of the normal performance model (intercept, slope), the

abnormal returns could be calculated. Afterwards, the accumulated abnormal returns were

estimated to get more insights in the impact of the particular event on the share price of Tesla

[3]. Eventually, it was tested whether the abnormal returns of Tesla were significantly

different from the expected returns 10 min prior and 10 min after the events occurred by

conducting two-tailed t-tests. The results for the two event studies are presented together with

the results of the qualitative text analysis in the results section.

The two authors of this study read and coded the news items of the seven top financial

the news item (e.g. direct quotations by Elon Musk, financial analysts or reporters),

information and prospects about Tesla’s share price, the main conclusion of the news item,

the prevalent sentiment of the news items and other remarks. The two coders discussed the

Intraday Tesla share

prices on

August 23, 2016

started. The sample of news items was divided into two sets so that every coder was able to

provide in-depth notations for the respective news items. After sorting all news items based

on the date and time of release, a clear timeline of the news reporting emerged that

facilitated the investigation of the news dynamics across outlets over time.

On August 23, 2016 at 11.23 a.m. EST, Elon Musk, the CEO of Tesla, announced on Twitter

that there would be a “product announcement at noon California time today.” Not only did

this tweet generate considerable engagement on Twitter (i.e. 5,251 retweets; 14,330 likes), it

also caused the Tesla share price to jump 1.4 percent from $223.14 at 11.23 a.m. to 226.37 at

11.30 a.m., with a trading volume of 549,981 shares in that period, which accounted for more

EST on the tweet and the rising share price of Tesla therewith: “rose more than 1 percent

after the post and were recently trading up $3.69.”

The results of the event study in which this first tweet by Elon Musk was treated as an

event showed that there was indeed a significant market reaction of the abnormal returns of

Tesla at the moment when the tweet was released (11.23 a.m. EST). Besides a few minutes,

the significant impact of the event on the Tesla share price persisted up to 9 min after the

tweet was released (see Table I and Figure 3). No significant abnormal returns could be

found for the 10 min before the tweet was published. However, the tweet did not have a

lasting effect on the Tesla share price. Already 30 min after the tweet was published, Tesla

stocks declined 0.4 percent down to $223.37, slightly higher than before the announcement.

the Najarian Family Office Co-founder, saw some unusual activity. He contemplated that

there might have been some investors who had access to inside information. When looking

Event 1: Tweet by Elon Musk (11.23 a.m.) Event 2: Announcement by Tesla (3.30 p.m.)

Time Abnormal returns Time Abnormal returns

11.23 a.m. 8,553*** 3.30 p.m. −13,263***

11.24 a.m. 9,860*** 3.31 p.m. −3,982***

11.25 a.m. 3,322* 3.33 p.m. −3,293***

11.27 a.m. 2,133* 3.34 p.m. −3,877***

11.28 a.m. 1,984* 3.35 p.m. 3,314**

11.31 a.m. −2,930** 3.36 p.m. 3,022**

event studies for tesla

(see Figure 4), it indeed seems that there might have been some preliminary trading on Tesla

shares before Musk tweeted about the upcoming product introduction. Later that day, Elon

Musk sent another tweet at 3:02 pm, advertising the revelation, “Journalists Q&A for

30 mins and embargo ends at 12.30.” However, Tesla shares reacted this time only mildly

(0.6 percent), rising from $226.16 at the minute of the release to $227.48, half an hour later.

Before the actual product announcement by Tesla at noon California time, there were two

more news items published. Bloomberg reported shortly on the reactions of the Tesla share

price to the tweet by Musk (“Tesla jumped 2 percent to $227.45 at 11:53 am after rising as

much as 2.3%, the biggest intraday gain in two weeks”) and referred negatively to the

investigation with regard to the Autopilot car crash as well as to the “trimmed” full-year

Intraday event study

announcement of the new product, Tesla’s stock was up at $228.09, 1.7 percent higher than

its opening price that day.