The stock tips from your buddy? They really aren’t good
Every investment professional starts to roll her eyes when a private investor starts to give her investment tips. We all know instinctively that these tips aren’t going to work out, but we never really had the data to prove it. But now, I can show you the performance of stock tips shared on a prominent investment website.
Hailiang Chen and Byoung-Hyoun Hwang examined server data from the prominent US website Seeking Alpha and checked which articles were most often read and which ones were most often shared with other users. To my surprise, they found that articles with a lot of numbers in them were read through less often but shared more often by the readers that made it all the way through it. Initially, I thought that should be a good sign because articles with a lot of numbers in them may be better researched and better underpin the case they make.
Until you realise that most articles on a site like Seeking Alpha that use a lot of numbers are simply reports on the most recent earnings release, etc. Thus, articles with a lot of numbers are predominantly backward looking and readers who share them don’t share useful information for future returns. Instead, they share numbers that make them look good in hindsight by showing what great stocks they picked and how well the companies they picked do.
I have my issues with the interpretation of the authors of why people share articles and what their intrinsic motivation to do so is, so I won’t go into that here. But what they nicely show is how someone would perform on average if she had taken these stock tips to heart and invested in the stocks when they received an article shared with them. The chart below uses two different ways to measure the virality of an article by comparing the number of shares with the number of times the article has been opened or read to the end. And it shows performance either as average performance or adjusted for past returns, market cap, and valuation (DGTW-adjusted). No matter how you slice and dice it, the stocks on average rallied for another 5 trading days before turning south and accumulating losses for the people foolish enough following their buddy’s stock tip.
Performance of stocks shared in stock tips
Source: Chen and Hwang (2021)
To my surprise, they found that articles with a lot of numbers in them were read through less often but shared more often by the readers that made it all the way through it. Initially, I thought that should be a good sign because articles with a lot of numbers in them may be better researched and better underpin the case they make. I keep reading latest updates on tejkohli.co.uk to make sure I am on right track. Thanks for sharing this article.
Joachim, appreciate the insight and I find this particular research interesting, mainly because I was just solicited by the director of Seeking Alpha to become a contributing member via a Twitter exchange. I tracked the link and the data the author is pulling is from 2012-2013. Obviously a lot has changed since then. I've also heard that SA will decline articles that are just a recap of earnings. Other people I know who have contributed at SA in 2013 had earning recaps declined because their focus wasn't forward-looking enough. Whatever that means.