The world of social media has evolved rapidly in recent years. Traders and investors widely use various platforms both as a means of getting access to news and for trading ideas.
Using social media as a trading indicator isn’t a new concept and the sentiment of tweets from Twitter has previously been shown to lead to outperformance in various different studies.
However, a new paper by Rakowski, Shirley and Stark, suggests that there is a significant edge to be had by moving beyond news ‘sentiment’ and focusing on how a news source such as Twitter, helps stocks gain increased levels of attention.
Is All That Twitters Gold?
Rakowski, Shirley and Stark focus on how stocks that have previously received limited exposure through either the mainstream financial media or analysts at major institutions, have the ability to sharply increase in value in the days following a spike in the number of Twitter posts they feature in.
The authors use hashtags (or more specifically cashtags – which are the tickers that use a $ sign) to identify stocks and they calculate the daily numbers of tweets they receive.
Between 2011 and 2015, the authors monitored the increases and decreases in daily cashtag activity for stocks that make up the Russell 3000 index at that particular point in time.
Stocks were also analyzed based on market cap, the number of analysts covering them, and the number of daily news articles.
A Long/Short Twitter Strategy
In the paper, the authors construct a long/short trading strategy. They take the top decile of stocks that have shown an increased level of Twitter activity and also those that have the lowest market capitalization.
In other words, they are looking for small cap stocks with high Twitter activity.
Holding the highest ranked stocks for one day after the spike in Tweets, resulted in an adjusted excess annual return of 58.10% or about 291% over the five year sample period.
Interestingly, the second day after a spike in Twitter activity, showed significant levels of mean reversion, whereby those same stocks gave up much of their gains.
There are a few consideration that we need to take into account with this study.
Firstly, it’s often those stocks that have low market caps and are the least liquid and lightly traded that are prone to the largest price increases because of a Twitter spike.
The results don’t include commissions or fees associated with executing trades. Instead, the authors conduct the strategy at the bid/ask midpoint to help model transaction costs.
The authors of the paper theorize that the reason for the level of outperformance comes down to what they call “Twitter attention”, which is a combination of investor attention and impediments to trade.
Investor attention refers to the fact that smaller, less liquid stocks, simply don’t get the level of attention or coverage that larger companies do and are therefore more likely to be mispriced.
Similarly, impediments to trade refers to the fact that small and illiquid stocks are difficult for large institutions to trade and are therefore more likely to be incorrectly valued by the market.
Key Takeaways for Traders
Even though many would suggest that Twitter is just a distraction, it appears that there is value in using it as a tool to identify stocks that have the potential to outperform.
If a low-cap stock that has limited analyst exposure, begins to show a significant increase in the number of tweets it’s receiving, then there is the possibility that it could outperform the day following that spike.
However, it’s important to note that those returns will likely be mean reverting the day following any price increase.
This information is particularly valuable for intraday momentum traders who are on the lookout for stocks that have the potential to make large intraday moves.
Deciphering the information content within the tweets (i.e. whether the sentiment is bullish or bearish) could result in even bigger gains for traders.
How To Get Ahead
There are essentially two ways that traders can use the information contained within this study to find an edge and get ahead.
The first is to built a mechanical trading strategy using a platform like Amibroker, Quantopian, Multicharts, or even Excel.
This method would require streaming data from Twitter as well as price data for Russell 3000 stocks.
The second way is to build a strategy of your own in real time.
To do so, start with a paper trading account and look out for high Twitter activity in the small cap space. Track all of your trades and evaluate your progress diligently.
If you are in profit after a sample of 30 trades, you may move to a small pot of real money. If you are happy after 100 trades, you can start to increase size.
This is a process I use to take new trading ideas into the live market and what I talk about in the System Traders Feedback Loop.
Keep analysing your progress to see whether Twitter is giving you an edge.
If you are on Twitter, you can follow me @marwoodjb.