Musk Effect: Social Media Influence on Share Price

This content is published on behalf of the Economic Policy Research Group

By Dain Rohtla 

Finance is an ever-changing field that shifts as the world around it does. With the growing presence of social media, it is indispensable in business. Behavioral finance takes people’s biases and judgement into account and assumes people do not always act in the most rational or logical way. As a young person myself, I wanted to blend my knowledge of social media and finance and see what links I could find. I researched to what degree do celebrities’ social media influence impact companies’ stock value. This came in two parts: 

1) Individual cases of celebrities posting about a specific company and determining the outcome. 

2) Assessing the correlation between a CEO or important figurehead and the stock value of their affiliated company. 

The purpose of these studies is to solve how this behavioral information can be used to generate alpha in the marketplace or be ingrained into the company’s internal strategy. 

Rogue Tweets 

Some social media content has broken the internet. In 2018, American adults were spending over 11 hours per day interacting with media. 254 of those minutes were with internet connected device or computer, and App/Web on a smartphone or tablet (Fottrell, 2018). This amount of exposure means posts can make big waves. The authority bias shows that people believe “Authority” figures regardless of the context and their expertise. This was proven by Milgram’s 1961 experiment. Social media is a powerful way to express authority because it is so prevalent in young people’s lives today and the explicit count of ‘followers’ quantifies power. This authority is leveraged to move markets. 

Public opinion has a strong effect on an individual’s thoughts. The idea of capturing the public’s mood is not new. Bollen, Mao, Zeng (2011) research was the creation of an algorithm to determine the direction of the DJIA each day. They had a success rate of 87.6%, the probability of this occurring by random chance is only 3.4%. This was used by the hedge fund, Derwent Capital Markets. Surprisingly, they were a failure and shut down a year after opening their doors and auctioned off their assets for a big loss (Lachanski and Pav, 2017). 

This leads us to believe you cannot trust the aggregate mood; Maybe only a few select individuals and institutions are important? Some influencers have used social media to sort out their beef and caused giant stirs in the market. People put extreme trust in authority figures even irrationally. Here, as in Milgram’s experiment, people follow the directions of the ‘authority’ blindly. 

Listed below are some of the most impactful social media posts. 

                Elon Musk Taking Tesla Private Increase 11%i 
                Elon Musk Smoking Pot on Joe Rogan podcast Drop 9%ii 
            Kylie Jenner Said did not use snapchat anymore Drop 6% ($1.3 Bn)iii 
                Carl Icahn Apple undervalued Increase 5% iv 
                Muddy Waters Claimed ADNC investigation Drop 25%v 
                Muddy Waters Said Huishan Diary Fraudulent Drop 82% (longterm)vi 
                Trump  Boeing Air Force One expense Drop 1% ($1 Bn)vii 
                Trump Cut Military expenses (Lockheed Martin) Drop 5% ($4 Bn)viii 
                Trump Toyota building in Mexico Drop 3% ($8.2 Bn)ix 
                Rihanna Criticized Snapchat Drop 5%x 
            H. Clinton Condemn drug price gouging (IBB) Drop 5%xi 
                DJ Khaled Partnered with Weight Watcher Increase 7.3%xii 
                Oprah Partnered with Weight Watchers Increase over  100%xiii (few days) 
                Jeremy Jordyn Illness from Chipotle Drop 2.7%xiv 
                John McAfee Crypto coins Various xv 

Evidently, these occurrences are rarely positive. This could be because negative reviews are more authentic than their positive counterparts. Influencers are known to be paid to promote products and services and this therefore degrades their authenticity. Negative comments are more personal and genuine and create a larger effect. This power can open fissures and create greater volatility, leading to large losses such as Musk realized once he was sued for his tweet that originally raised his company’s value. 

Follower : Twitter Correlation 

My hypothesis for the second study was that well known CEOs would have a strong link between their twitter followers and their company’s stock value. I started by comparing two per sector, one widely known, and the other largely unknown. I used the Pearson Movement Correlation to find the Correlation between the percent change in Twitter followers to percent change in stock price.  

Computers: I started with computers and compared Lenovo and their CEO Yuanqing Yang to PC giant Windows and founder Bill Gates. Bill Gates is no longer Chairman or CEO, yet he still has a strong connection to the company in the minds of people. The current CEO is Satya Nadella, a much less well-known figure and has less than two million followers on twitter. I had less available data on Yang than the others, only from the start of November 2017. He had a startling correlation of –0.408924483 to his company’s value. He posts very infrequently which may play an effect, but also posts less regarding the business world, and is a clear fan of Scuderia Ferrari. Bill Gates had a greater correlation per my thesis, 0.033152. Data ranged from start of August 2016 until present. He does not post much about Microsoft and instead almost entirely about his philanthropy.  

EV: I was most excited about Elon Musk as his August tweet regarding taking Tesla private that led to share price surging 11% by the end of the day was what sparked my interest in this study. His wild antics are spread among tabloids and he is a beloved figure by young people not only for his business acumen, but his technical skill, lofty goals and eccentric behavior. His correlation was shockingly low, -0.11001. This may be because his fame is not tied to twitter followers but is more encompassing. 

Musk’s competitor in this electric car sector is Padmasree Warrior, former CEO of EV manufacturer Nio. They are most well known for their supercar, Nio EP9, one the fastest car in the world around the benchmark performance test track the Nürburgring.  This is the most tenuous relationship. Her data was not abundant, Twitter followers only counted from October 2018, and she left the company in November 2018, albeit she is still the most well-known figure at the company and is listed by Google as the CEO. Her correlation is –0.71605 and only has about 1.5 million followers. 

Social: The next comparison is flimsy, yet the figures are still useful. I compared Jack Dorsey, Twitter CEO as the unknown CEO to giant Donald Trump, supposed driver of the US economy. Jack Dorsey had a significant correlation of 0.12109927. Surprisingly his tweets about his company, Twitter, are almost nonexistent. He uses it more like a regular user does, random, non-economically influenced posts.  

Trump is well-known for his liberal use of the social media platform and is berated nearly daily on the news. He has used it unlike any politician before him and it changed the game. The start of 2018 had phenomenal growth in the stock market. 

Even more startling than this number is how the market reacted after he was elected. The S&P 500 did not have a single negative day between his election and inauguration. The market rose to a high of 6.4% during that period, a stark difference from past president Barack Obama. The S&P 500 hit a high of –5.3% during the same period and a low of –25.2%. xvi  Not surprisingly he has a great correlation to the overall market, 0.213355.  

E-Commerce: Bezos is the world’s richest man and with that title comes amazing fame. Surprisingly he lacks twitter followers and has less than 1 million. His correlation suffers and is represented with a measly -0.033. Devin Wenig CEO of Ebay is well known and is a Twitter novice. His account has around 11,000 followers, which probably is not much more than some college-girls you may know. His correlation is horrendous, -0.365.  

As you can see for each of these four comparisons, the greater the celebrity the higher the correlation. Yet some of these have insignificant levels, or even negative. The use of their account is important as well. Some use it as an extension of the company, the account is likely run by an assistant or the PR team. These tend to suffer in the correlation score. Musk, Dorsey, and Trump all use the platform brazenly and informally thus achieving the largest effect. Musk has a low correlation, but individual tweets have had huge effects. Trump and Dorsey have the two highest correlations and use it similarly. The secret sauce is to have a lot of follower base but also use it personally and not be an extension of the company. 

                  Individual                     Followers                  Correlation 
                      Trump                     59,369,410                  0.213 
                      Gates                     46,888,063                  0.033 
                      Musk                     25,472,563                  -0.110 
                      Dorsey                      4,171,994                  0.121 
                      Warrior                     1,458,287              -0.716 
                      Bezos                     878,392                  -0.033 
                      Yang                     53,926              -0.409 
                      Wenig                     11,775                  -0.365 

xvii 

Uses 

This knowledge is important in two capacities. The internal Public Relations teams need to understand the effects of social media for the company to succeed. Externally, it can be useful for investors seeking to generate Alpha. 

PR 

Understanding that success on social media relates to stock success is important. As followers grow so does stock value. Companies should try and build the presence of their CEO. This is not done only with content of their business but in exciting and inspirational ways.  

Investment Opportunity 

Public Sentiment is becoming a more mainstream way to value stock value. Ranco et al (2015) discuss this in their paper about how the volume and sentiment of tweets change stock prices.  This is easy with the organized use of “cashtags” to denote stock. Sentiment is not a fantastic predictor though as other variables must be considered rather than just volume. Authority should take a large role. NPR had the idea and created a trading algorithm ‘BOTUS‘ to monitor Trump’s trades. It ranked them neutral, positive, or negative based on what Trump said and then bought or short sold. It acted at the time of tweet and covered 30 minutes later. It started April 7, 2017 and shut down November 21, 2017. losing $0.37 or 0.037% (Goldmark, 2017).  On April 7, 2017 Market opened at 2,356.59 Closed Nov 21 at 2599.03 a positive change of 10.29%. This showed not to be effective strategy. While Trump’s tweets affect some instances, it is not enough to profit on. 

Wall Street Journal created an index of 12 stocks Trump tweeted about in the past year and tracked their value. At the time of the article it was up 32.5%, beating the S&P500 and DJIA (Ingram, 2017). All but only one company he mentioned rose in value (Toyota), even those he dislikes such as Nordstrom. Mentioning, regardless of sentiment could be more valuable. This again must be taken with a grain of salt because the market rose broadly during this time. 

Fame has been easier and quicker to garner with the rise of social media. With this, fakes can make their way into plain sight without being vented as easily. This is a key vulnerability to sentiment. Oobah Butler is the best example of creating something totally fake on the internet. Vice Reporter Oobah Butler published his story, I Made My Shed the Top-Rated Restaurant on TripAdvisor, on December 6, 2017. In it he reported how his restaurant became the highest rated eating establishment in London with nothing but social media, not even existing. Through raving reviews, it took the spot. His photos of the foods even mocked the hungry fans. One of his photos of a meal on the website was a painted sponge and shaving cream, another was an egg on his bare ankle. People looked through all rationality to believe solely on the reviews. They even neglected the fact that it existed in a very residential area and did not have an address, only a street name. Despite this people called and emailed incessantly to book. If public sentiment is misguided like this, it could lead to huge downturns. 

Inflated Value 

Below is a list of CEOs listed om descending Twitter follower counts. Larry Page, Warren Buffet, and Mark Zuckerberg do not have Twitter accounts. There placement can be argued but they are placed there due to the relative size of their companies and news-appeal. This chart shows how people with more followers generally have a higher cost ratio than the industry average and how that trends down as followers decrease. 

      CEO Twitter Followers       P/B        P/E Industry Av. P/B  Industry Av. P/E Multiple P/B Mult. P/E 
Trump 58,780,836       1.49      11.92    1.07  6.82      1.39  1.75 
Gates 46,680,129       9.38      26.17    8.38  30.11      1.12  0.87 
Musk 25,105,873       10.33      NA    1.62  13.29      6.38   NA 
Cook  11,078,425       6.06     17.67    1. 83  12.08      3.31  1.46 
Page NA       4.64     35.97    5.70  41.41      0.81  0.87 
Buffet NA       1.44     146.75    1.08  19.97      1.33  7.35 
Zuckerberg NA       5.51     21.41    5.31  33.91      1.04  0.63 
Legere 6,161,520       2.64     27.01    2.14  55.32      1.23  0.49 
Dorsey 4,144,189       2.78     19.69    5.70  33.91      0.49  0.58 
Bezos 868,279       18.85     84.29    6.22  48.86      3.03  1.73 
Yang 53,750       1.49     21.14    8.38  30.11      0.18  0.70 
Wenig a11,694       5.44     5.83    5.31  41.41      1.02  0.14 

xviii 

Criticisms and Advancements 

This paper leaves some areas for improvement. It Is worth noting that one could criticize Gates, Warrior, and Trump for correlation to their affiliated stock. In addition, not all these subjects had twitter for the same amount of time, some have not used it for long.  

There is room for advancements on this topic. Granger Causality Tests would be insightful to run, as would the correlation between number of tweets and stock price. 

Conclusion 

In the age of connectedness people have power like they never have before. It can come swiftly and to anyone. Influencers garner this power through platforms such as Instagram, Twitter and Snapchat. Users with large followers have an army behind them to change a company’s value and it can be changed in minutes. These are most often short lived. few, such as Oprah’s affiliation to Weight Watchers, have lasting effects.   

Correlation between follower movements and stock value generally rise with the number of followers but there is not enough data available for a conclusive result because few public Fortune 500 CEOs use social media. Follower trends is not a great indicator of stock movement, especially because both historically go up rather than down. 

Well-followed Chief Executives often leads to inflated financials. This could be due to people putting unfounded belief in these companies because their social media presence makes them more visible and relatable. 

Looking into behavioral effects on share value is more important for the firm itself rather than outsiders looking to develop alpha for their portfolio. If an outsider is looking to profit from social media sways, the changes are mostly intraday and not good predictors of long-term outlook. A large following enables a person to create large sudden price movements. This can also get them in hot-water such as Musk found out when the SEC cracked down on him for tweeting about taking Tesla private. A larger social media presence could lead to better interaction with consumers and lead to an increased share price. 

References 

Berg, M. (2018). The Rihanna Effect: Snapchat CEO Evan Spiegel’s Net Worth Drops Nearly $150 Million In Two Days. [online] Forbes. Available at: https://www.forbes.com/sites/maddieberg/2018/03/17/the-rihanna-effect-snapchat-ceo-evan-spiegels-net-worth-drops-nearly-150-million-in-two-days/#373cbfba2ca0 [Accessed 29 Mar. 2019]. 

Bollen, J., Mao, H. and Zeng, X. (2011). Twitter mood predicts the stock market. Journal of Computational Science, [online] 2(1), pp.1-8. Available at: https://www.sciencedirect.com/science/article/pii/S187775031100007X?via%3Dihub  [Accessed 29 Mar. 2019]. 

Butler, O. (2017). I Made My Shed the Top Rated Restaurant On TripAdvisor. [online] Vice. Available at: https://www.vice.com/en_uk/article/434gqw/i-made-my-shed-the-top-rated-restaurant-on-tripadvisor  [Accessed 29 Mar. 2019]. 

Egan, M. (2015). Hillary Clinton tweet crushes biotech stocks. [online] CNNMoney. Available at: https://money.cnn.com/2015/09/21/investing/hillary-clinton-biotech-price-gouging/index.html  [Accessed 29 Mar. 2019]. 

Fottrell, Q. (2018). People spend most of their waking hours staring at screens. [online] 

MarketWatch. Available at: https://www.marketwatch.com/story/people-are-spending-most-of-their-waking-hours-staring-at-screens-2018-08-01  [Accessed 29 Mar. 2019]. 

Giammona, C. (2018). Weight Watchers Shares Jump After DJ Khaled Endorses Program. [online] Bloomberg. Available at: https://www.bloomberg.com/news/articles/2018-01-02/weight-watchers-jumps-after-dj-khaled-joins-oprah-as-endorser  [Accessed 29 Mar. 2019]. 

Goldmark, A. (2017). Episode 763: BOTUS. [online] npr. Available at: https://www.npr.org/sections/money/2017/04/07/522897876/meet-botus-planet-money-s-stock-trading-twitter-bot  [Accessed 29 Mar. 2019]. 

Ingram, M. (2017). http://fortune.com. [online] Fortune. Available at: http://fortune.com/2017/02/24/trump-tweet-stocks/  [Accessed 29 Mar. 2019]. 

Lachanski, M. and Pav, S. (2017). Shy of the Character Limit: “Twitter Mood Predicts the Stock Market” Revisited. Econ Journal Watch, [online] 14(3), pp.302-345. Available at: https://risk.princeton.edu/img/Lachanski2017.pdf  [Accessed 29 Mar. 2019]. 

Ranco, G., Aleksovski, D., Caldarelli, G., Grčar, M. and Mozetič, I. (2015). The Effects of Twitter Sentiment on Stock Price Returns. PLoS One, [online] 10(9), p.e0138441. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4577113/

Randewich, N. (2018). Two years in, Trump holds stock market bragging rights. [online] 

Revesz, R. (2016). Donald Trump wipes off $1bn from Boeing’s stock market value with single tweet. [online] The Independent. Available at: https://www.independent.co.uk/news/business/donald-trump-targets-boeing-shares-fall-twitter-air-force-one-planes-dennis-muilenburg-a7458946.html  [Accessed 29 Mar. 2019]. 

Salinas, S. (2018). Tesla stock closes down 6% after top executives resign and Elon Musk smokes weed on video. CNBC. Available at: https://www.cnbc.com/2018/09/07/tesla-sinks-8percent-after-bizarre-musk-podcast-appearance-cao-exit.html  [Accessed March 29, 201 

The Sydney Morning Herald. (2017). Toyota shares drop after targeted by Trump over Mexico manufacturing. [online] Available at: https://www.smh.com.au/business/toyota-shares-drop-after-targeted-by-trump-over-mexico-manufacturing-20170106-gtmza1.html  [Accessed 29 Mar. 2019]. 

Thielman, S. (2016). Trump’s tweet about Lockheed-Martin cuts $4bn in value as share prices fall. [online] the Guardian. Available at: https://www.theguardian.com/business/2016/dec/12/lockheed-martin-share-prices-donald-trump-tweet  [Accessed 29 Mar. 2019]. 

Tsukayama, H. (2013). Apple shares jump on tweets from Carl Icahn. [online] The Washington 

Post. Available at: https://www.washingtonpost.com/business/technology/apple-shares-jump-on-tweets-from-carl-icahn/2013/08/13/b2c32ffe-044a-11e3-88d6-d5795fab4637_story.html?noredirect=on&utm_term=.42e9b932c058  [Accessed 29 Mar. 2019]. 

Ren, S. (2017). Huishan Dairy Crashes 90%: Chairman Refutes Allegations. Barron’s. Available at: https://www.barrons.com/articles/margin-call-huishan-dairy-tumbles-90-muddy-waters-cheer-1490328082  [Accessed March 29, 2019]. 

Whitten, S. (2017). Chipotle shares tank after actor Jeremy Jordan said he ‘almost died’ after eating at the chain. CNBC. Available at: https://www.cnbc.com/2017/11/13/actor-jeremy-jordan-says-he-almost-died-after-eating-at-chipotle.html  [Accessed March 29, 2019]. 

All PB and PE ratio and industry average data comes from Reuters financials. 

All follower data comes from socialblade.com 

All Stock prices come from Nasdaq.com Historical Quotes 

Image Source: Frederic J. Brown/AFP/Getty Images

The opinions expressed in this article are the author’s own, and may not reflect the views of The St Andrews Economist

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s