One of the overriding questions from the 2016 presidential campaign is whether the onslaught of “fake news” altered the outcome of the election. That question is unanswerable, but we now know whether the financial equivalent of fake news altered stock prices.
In early 2017, the SEC took action against a number of web sites, including Seeking Alpha, after it uncovered a large number of articles that contained false information and were posted to prop up the stock prices of the underlying companies. A just-published research study, Fake News, Investor Attention, and Market Reaction, showed that those articles succeeding in driving traffic and raising awareness of the companies.
But it showed that investors were smart enough to disregard the information. There was no discernable increase in the share prices of the companies.
The immediate takeaway is that the market was sufficiently efficient and investors were adequately discerning so that this type of fake news did not pose a threat to you or your clients’ assets. But, as I will explain, there are other lessons about whether one can rely on crowdsourced information sites, such as Seeking Alpha, for accurate information.
First, however, let’s review what led to the SEC’s actions and what the researchers found.
C-list actress turned equity analyst
Kamilla Bjorlin is an actress from Encino, California, who performs under the name Milla Bjorn. She has had minor parts in a few obscure movies. But as a second career, she ran Lidingo Holdings LLC, which, over a two-and-a-half year period between 2011 and 2014 paid writers to generate hundreds of stories extolling the virtues of companies such as Advanced Medical Isotope Corp. (ticker: ADMD) and Galena Biopharma, Inc. (ticker: GALE).
Public companies are allowed to use, for example, public relations firms to promote their interests. But in this case, Bjorlin (who wrote many of the articles herself) and her writers failed to disclose that they were being paid by the company.
This came to light in April 2017, when the SEC announced a settlement with Bjorlin’s firm and 16 others for engaging in deceptive stock promotion. The cases were settled for amounts ranging from $2,200 to $4.8 million.
The settlement cited roughly 450 articles that appeared on a number of sites, including Benzinga, Forbes, Investor Village, Minyanville, Small Cap Network, The Motley Fool, TheStreet and Yahoo Finance. But, by far, the greatest number appeared on Seeking Alpha.
None of the web sites were charged by the SEC – only the authors, their employers and some of the companies that paid for their services. Indeed, the complaint did not state that the sites were aware that the authors were paid promoters. It was the failure to disclose their affiliation with the underlying companies that broke the law.
Four researchers – Jonathan Clarke, Hailiang Chen, Ding Du, and Yu (Jeffrey) Hu – authored the above study. They were provided data by Seeking Alpha and studied only the stories posted on that site.
Fake news generated a lot more readership than legitimate news (about 82% more page views), according to the authors. Moreover, visitors to the site could not distinguish what was real from what was fake. There was no statistically significant difference between the number of comments on real and fake articles, and the comments on the fake articles did not contradict the article any more than in a control sample.
Those findings are illustrated in the following graphs, which appeared in the study. They show that fake news attracted more attention than real news, based on the number of page views, unique visitors, whether the article was read to the end and whether the comments on the article were read.
In fairness, the fake news articles were a very small fraction of the content published on Seeking Alpha.
The editors at Seeking Alpha score its content based on how convincing, actionable and well-presented each article is. The fake articles scored as well as the real ones, so even the editors were not able to identify the fake news.
Most importantly, the authors found that the stock market discounted the fake news. According to the authors, “the stock price reaction to fake news articles is not statistically different than a matched control sample over both short and long-run windows surrounding the release of the fake news.”
The average market capitalization of the companies with fake news articles was $41.3 million, which is in the micro-cap category and approximately 1,000 times smaller than the average size of companies with legitimate articles.
The issue is trust
It’s easy to dismiss this episode because it had no measurable impact on investors’ portfolios. But that ignores the larger issue of whether web sites upon which we rely are being systematically infected with unreliable content.
The problem of misinformation in social media is not confined to financial data. Last week, the New York Times reported on college students who are being paid by manufacturers to promote products – from headphones to wine to clothing – by posting pictures on Instagram and other social media sites. In some cases, the students had never used the products themselves. The FTC requires social media users disclose their affiliation with the manufacturer, but according to the Times article those guidelines are “often ignored.”
The problem with Seeking Alpha and the other social media sites is not their crowdsourced architecture. There are many examples of crowdsourced platforms that have succeeded in providing a valuable service. For example, a platform called GlobalXplorer facilitates collaboration among users to identify archeological sites.
Nor is it reasonable to expect the problem to be resolved by regulation or by more aggressive moderation by the firms that maintain the social media sites. The potential profit from a campaign that succeeds in raising a company’s stock price is too great.
The temptation to earn those profits is driven by the anonymity of the user. As long as someone can post to a social media site anonymously or without having their identity verified, the potential for abuse will persist.
Don’t rely on sites that permit anonymity for anything more than entertainment value. Those include not just social media sites like Seeking Alpha, but discussion boards where readers can comment on articles. Unless the site has a procedure for authenticating the identity of its members and those members are posting under their real names, anything you read should be carefully verified.
Read more articles by Robert Huebscher
