One of the most important platforms people use to vent out their feelings is social media.
Even when it comes to trading, a lot of people like to voice their opinion or tell what they are feeling towards the market. Some do it so that they can get some valid explanation or justification as to why the market is behaving this way or some do it just because they are used to telling it out loud.
This is what makes social media platforms like Twitter so powerful. Social media can be mined for data that reveals what people are thinking and feeling. And that, in turn, could translate into powerful investment ideas.
Twitter: A powerful tool for stock trading
Twitter is particularly useful. Usage of social media like Twitter has exploded in recent years, giving analysts a real-time reflection of popular sentiment. There are close to 400 million tweets every day and they range from a variety of topics, which also includes the stock market.
Apart from that, it also covers a wide variety of moods, i.e. people actually express how they are feeling towards a particular issue through their tweets. A lot of Institutional Investors, online brokers and retail investors are very active on Twitter, providing their valuable information through hash tags.
Financial Markets and Human Sentiments
A basic premise of behavioral economics is that the markets aren’t perfectly rational machines, but are expressions of human emotions like greed and fear. Twitter is one of the greatest tools ever invented for an immediate gauge of those human sentiments. This fact has been used by a lot of analysts to come up with a brilliant idea.
Analysts have invented tools that can use these millions of twitter feeds to filter topics and then judge the emotions of each individual towards a particular subject. This approach has been used pretty effectively to monitor the market and predict market changes based on the Twitter feeds.
One has to remember that there is a huge amount of nonsense on twitter and tweets are flooded with useless bicker. So filters have to be specifically designed to cater to one’s requirements. Analysts use a number of indicators to analyze these tweets and provide traders with alerts on events. They pick up socially expressed emotions on Twitter in real time and then use them to analyze and predict the market, something which is not available on traditional news channels.
The difficult part is not data collection, but data crunching. For example, there might be some 400 million tweets generated every day. From that, we have to first filter those tweets which are useful/ important for us. Once that is done, the next big hurdle is studying those keywords to predict the mood of the people through their tweets.
This is a recursive approach and would require analysis at different levels. One has to consider hundreds of factors while analyzing these tweets, then sort data according to the requirements and then provide results based on the mood of the public towards a particular instrument. Then there are lots of ways these generated results can be used. Analysts don’t always go with the general public opinion and some market analysis is necessary before a final result.
Social Media means Real Time Data
The biggest benefit of using social media is that the data sets on which one is working is immediate. It is much better than taking a survey every week. The data one is working on is genuine and is personal. People are not hesitant to say what they want to say. The social-media sentiment analysis is immediate and outgoing. One has to accept that the markets are moved by emotion. Many analysts think this is going to be the future of trading. One can actually see global moods moving up and down in real time.
In an interesting paper in The Journal of Computational Science called “Twitter moods predicts the stock market”, authors Johan Bollen, Huina Mao and Xiaojun Zeng start by arguing, “Behavioral economics tells us that emotions can profoundly affect individual behavior and decision making.”
They wanted to find if “Twitter feeds are correlated to the value of the Dow Jones Industrial Average (DJIA) over time.”
They found that by including the effects of public moods and opinions, the accuracy of DJIA predictions can be significantly improved. They found an accuracy of about 86.7% while predicting the changes in the closing values of the DJIA.
What we think
All this said and done, it is actually not that easy to dive into this domain as a private investor and start analyzing Twitter Data to make investment decisions. It is easier for Institutional Investors to start because they have their own proven models to predict the market and they can thus test this twitter driven model to see how effective it is.
They can tweak the new model in ways so that it matches and behaves according to their already existing model. Apart from that, given the amount of data we are looking at it is very difficult for individual investors to think they can make any sense of the 100 million active Twitter users.
Authentication of data is another problem as there are a lot of users who aren’t even verified and therefore your results might be disturbed by the wrong information from these unconfirmed sources. So you must be experienced enough and should have the tools to leave out such information or account for them.
But on the whole, there have been cases where Twitter-driven data has been effective in the predicting of particular stocks on the stock market and this tool has a lot of potential if explored well. There were also case studies as how Twitter-driven models were actually able to predict the disappointing debut of Facebook on the Stock Exchange. We thus have to realize how social media is not just a platform to vent and share, but a platform to learn, analyze and understand as well.