The popular small investor online forum WallStreetBets (WSB) had been on average producing sound investment advice, a team of researchers concluded.
The Reddit forum, now sporting over 10 million subscribers, gained spotlight after many of its users organized to buy GameStop stock earlier this year, starting a buying spree that boosted stock price of the video game retailer tenfold. The impact has drawn attention of establishment market players, lawmakers, and regulators.
Contrary to some criticism of the forum users as unsophisticated, if one followed advice of stock analysis posts on the forum in the past few years, one would have made significant returns, found researchers from the University of South Florida (USF) and University of Kentucky. They noted the forum has massively expanded in recent months—a change their initial analysis doesn’t capture.
The team looked at posts tagged as market research or “due diligence” (DD) by the users. They only looked at the ones that gave clear advice to buy or sell an individual stock between 2017 and the end of 2020, before the GameStop spree took off.
If an investor blindly followed the advice from the roughly 1,900 buy recommendations, he or she would have made more than by investing into a broad stock portfolio tracking the S&P 500 benchmark index. The recommended stocks yielded returns on average 1.2 percent higher than S&P 500 over two days, nearly 4 percent higher over 3 weeks, and more than 6 percent higher over 13 weeks.
The numbers are notable, the authors said.
“To consistently outperform the market you need to have some good information source or good trading strategy that is not yet completely incorporated by the market because if everybody knew about it you wouldn’t be able to earn this return,” said Jan Hanousek, USF PhD student and co-author of the paper.
By comparison, following recommendations of the Seeking Alpha crowd-sourced financial analysis site yielded returns of just 0.4 percent above overall market performance over three days, according to a different paper published last year.
Hanousek said the quality of the WSB recommendations varied a lot.
“There are posts that are much better and that can give you like 3, 5 percent return for the month, there are some that are worse and that will give you zero,” or even result in a loss, he told The Epoch Times.
That also means users could benefit from being able to discern higher quality advice, he said.
“If you are a skilled trader and you can actually make good inferences from the different analyses and make your own hypothesis … you can actually make even larger return [than average].”
The authors checked if there was a way to spot the highest quality DD posts, by, for example, pinpointing the user accounts that most likely post them. They couldn’t find such a pattern. Usually, one user account would produce only a single DD post.
It’s common for people on Reddit to use multiple accounts, Hanousek pointed out, so it’s hard to tell if there were some skilled analysts who were protecting their identity by switching accounts between DD posts.
The researchers also found that each DD post was followed on average by about a 7 percent increase in retail trading in the recommended stock. That raised the question of whether the stocks increased in price solely because of the buying pressure from people responding to the recommendations.
The paper explored that question, concluding that in that case the price hike should have reversed as the buying pressure subsided. But the fact that the positive returns endured over weeks and even months indicates that the DD posts were on average genuinely predictive and not just self-fulfilling prophecies, according to the authors.
Moreover, they found that the retail trading activity increased more after recommendations that yielded better returns.
“This pattern is consistent with retail investors being able to discern DD report quality,” the paper said. “It also casts doubt on regulators concerns that WSB is inducing retail investors to make uninformed trades.”
The investment returns following the DD posts were better for larger retail investors, said another co-author and PhD student at USF, Zicheng Xiao. The team is looking further into this point.
“It seems very reasonable that traders with larger capital for trading have better skill and can identify the better DD posts,” Hanousek commented via email.
The authors are now working on a revised version of the paper. One additional analysis, conducted by co-author Russell Jame, associate professor of finance at University of Kentucky, shows that the DD posts predict changes in the recommended company fundamentals.
“They are predicting that something good is going to happen for the company and it’s essentially discovered by the users on WallStreetBets,” Hanousek said.
Jame measured the phenomenon by looking at several indicators of sentiment regarding company cash flow over 5 and 21 days after a DD post.
“We find strong evidence that Buy DD posts predict cash flow news,” Jame told The Epoch Times via email.
When it came to “sell” recommendations on WSB, there were only about 200 posts over three years. Following them would have lost one about one percent on average over 5 days and “close to zero regardless of the holding period,” the paper said (pdf).
“While the estimates are not small, … they are not reliably different from zero,” likely due to the small sample size, it said.
Both Hanousek and Bradley said they expected this result. Sell recommendations are by nature less attractive to small investors who are less likely to hold the stock in the first place. For this reason, the recommendation to “hold,” which is common among establishment analysts, wasn’t even included in the paper. For a small investor, “hold” is often the equivalent to “buy,” Hanousek noted.
The team started to work on the paper months before the GameStop rush. The idea came from a conversation Hanousek had last summer with Xiao, who worked on a study of the quality of investment advice given by YouTube personalities. Hanousek, a WSB follower himself, recommended the forum as subject matter instead as its text form was better suited to data collection and analysis. Bradley, their teacher at the time, then came onboard followed by Jame.
Xiao told The Epoch Times he decided to study investment advice on social media following the advent of online trading apps in recent years that provided investment access to the broader public, particularly younger generations.
“I find more and more people, they try to obtain some investment opinion on social media, but they do not have long-term investment experience,” he said. “I was worried about that maybe they do not have the ability to tell if the news is true or not.”
The team is currently working on incorporating data from the first quarter of 2021 into their analysis, which they plan to release in June.
They noted that the WSB forum has dramatically increased in size and exposure since the GameStop rush, which may have changed the quality, and certainly the quantity, of DD posts.
“How these changes impact both the content and value of the investment advice on WSB is an interesting and fruitful area for future research,” the paper says.
Be the first to comment