Twitter Inc, a website and mobile platform where individuals post a 140 (and soon to be 280) character thought, or “tweet”, represents one of the newer social media resources out there. It was only a matter of time before individuals started to use it to develop stock and market trading strategies and they may be on to something.
A 2012 study by researchers at UC Riverside and Yahoo! Research Barcelona concluded that analyzing tweets about the share prices of underlying companies could improve stock trading strategies. The study, entitled “Correlating Financial Time Series with Micro-Blogging Activity” considered Twitter the foremost micro-blogging authority and created sets of filters to analyze tweets about companies and whether this helped improve trading returns.
A measure of tweeting activity looked at total tweet volume, hashtags about specific firms or subjects, as well as using a dollar sign and ticker symbol, which denotes an individual stock on Twitter. The general mention of a company’s products or its financial position were also considered. Retweet activity where a user tweets an original tweet to his or her followers, was also looked at.
The basic premise of the study was to look for a correlation between the above items and share price performance. The conclusion was that there is some predictive value on tweet activity and share price performance. In particular, the authors found a stronger correlation of Twitter action and share price performance for firms with low levels of debt, stocks with high betas, or a low float. Generally, the study stated that firms with financial performance that fluctuated significantly had a high correlation of Twitter mentions and trading volume, which seems to indicate the opportunity to trade on volatility and either invest in a stock or go short to profit from a fall in the stock price.
Hedge Fund Twitter Strategy
A recent Bloomberg article detailed that hedge funds are increasingly implementing trading strategies based off information they glean from Twitter. One fund uses social media feeds from Twitter, as well as related platforms such as Facebook Inc. to gauge investor sentiment and try to profit from either positive or negative stock market sentiment.
The use of algorithms is not surprising given this is a set of rules developed for tracking sentiment and creating trading strategies to act on news flow. Filters have also been mentioned as a way to analyze data and find meaningful trends, as opposed to noise with no real trading potential.
An estimated 400 million tweets are created on Twitter every day, which means the need to parse the information to more basic components. One strategy detailed eliminating everything but five important tweets for a company. Another one detailed hunting for tweets that have the potential to move the market, or an individual stock, significantly. Whether this information turned out to be true was beside the point. One faulty tweet supposedly stemmed from the White House and falsely detailed an explosion in the White House. Regardless, the market moved down significantly on the news, though it proved to be only short-lived.
Rise of the Machines
The opinions of academics and hedge funds suggest potential in Twitter-based trading strategies. Fortunately for individual investors, there are data-mining resources available for anyone to try and study tweets for trading opportunities. One such provider is Dataminr and another, data and analytics firm, Knowsis of the United Kingdom was built on the idea that social media activity could move the markets and individual share prices. They call themselves "a web intelligence company" that extracts "value from non-traditional online sources into quantifiable and actionable output for the capital markets."
Public sentiment is always an important short-term driver of asset prices, and social media can prove to be the canary in the coal mine (or for more positive news flow) that individual investors can catch just as easily as larger institutions and hedge funds. In a sort of sense, social media can level the playing field across investors of varying sizes. Even stock exchanges have started to analyze Twitter data to try and discern whether there is predictive value on share price volume from tweets. This could help them better prepare for trading volume rushes, or volatility that may trip trading volume circuit breakers.
In Summary
The Internet and digital information have been ripe for analysis and an attempt to use data to profit with stock-trading strategies. Social media represents the most recent incarnation of electronic data. Since it is in its infancy, the potential is probably greatest to exploit inefficiencies, be it data miners that wish to sell their data, or individuals and institutions that want to make money from it.
Ryan Fuhrmann can be contacted at Rational Analyst
A 2012 study by researchers at UC Riverside and Yahoo! Research Barcelona concluded that analyzing tweets about the share prices of underlying companies could improve stock trading strategies. The study, entitled “Correlating Financial Time Series with Micro-Blogging Activity” considered Twitter the foremost micro-blogging authority and created sets of filters to analyze tweets about companies and whether this helped improve trading returns.
A measure of tweeting activity looked at total tweet volume, hashtags about specific firms or subjects, as well as using a dollar sign and ticker symbol, which denotes an individual stock on Twitter. The general mention of a company’s products or its financial position were also considered. Retweet activity where a user tweets an original tweet to his or her followers, was also looked at.
The basic premise of the study was to look for a correlation between the above items and share price performance. The conclusion was that there is some predictive value on tweet activity and share price performance. In particular, the authors found a stronger correlation of Twitter action and share price performance for firms with low levels of debt, stocks with high betas, or a low float. Generally, the study stated that firms with financial performance that fluctuated significantly had a high correlation of Twitter mentions and trading volume, which seems to indicate the opportunity to trade on volatility and either invest in a stock or go short to profit from a fall in the stock price.
Hedge Fund Twitter Strategy
A recent Bloomberg article detailed that hedge funds are increasingly implementing trading strategies based off information they glean from Twitter. One fund uses social media feeds from Twitter, as well as related platforms such as Facebook Inc. to gauge investor sentiment and try to profit from either positive or negative stock market sentiment.
The use of algorithms is not surprising given this is a set of rules developed for tracking sentiment and creating trading strategies to act on news flow. Filters have also been mentioned as a way to analyze data and find meaningful trends, as opposed to noise with no real trading potential.
An estimated 400 million tweets are created on Twitter every day, which means the need to parse the information to more basic components. One strategy detailed eliminating everything but five important tweets for a company. Another one detailed hunting for tweets that have the potential to move the market, or an individual stock, significantly. Whether this information turned out to be true was beside the point. One faulty tweet supposedly stemmed from the White House and falsely detailed an explosion in the White House. Regardless, the market moved down significantly on the news, though it proved to be only short-lived.
Rise of the Machines
The opinions of academics and hedge funds suggest potential in Twitter-based trading strategies. Fortunately for individual investors, there are data-mining resources available for anyone to try and study tweets for trading opportunities. One such provider is Dataminr and another, data and analytics firm, Knowsis of the United Kingdom was built on the idea that social media activity could move the markets and individual share prices. They call themselves "a web intelligence company" that extracts "value from non-traditional online sources into quantifiable and actionable output for the capital markets."
Public sentiment is always an important short-term driver of asset prices, and social media can prove to be the canary in the coal mine (or for more positive news flow) that individual investors can catch just as easily as larger institutions and hedge funds. In a sort of sense, social media can level the playing field across investors of varying sizes. Even stock exchanges have started to analyze Twitter data to try and discern whether there is predictive value on share price volume from tweets. This could help them better prepare for trading volume rushes, or volatility that may trip trading volume circuit breakers.
In Summary
The Internet and digital information have been ripe for analysis and an attempt to use data to profit with stock-trading strategies. Social media represents the most recent incarnation of electronic data. Since it is in its infancy, the potential is probably greatest to exploit inefficiencies, be it data miners that wish to sell their data, or individuals and institutions that want to make money from it.
Ryan Fuhrmann can be contacted at Rational Analyst
Last edited by a moderator:
