Summary: A new hedge fund will be using 'sentiment analysis' of Twitter, a real-time social networking communication tool, to predict changes in the stock market. By using software that they have developed, they say that they have already been able to successfully predict profitable trades.
It's very interesting that someone is finally using Twitter to make financial decisions. As technologies like Twitter allow for faster and faster communication, they become more useful in financial decision-making. This is because changes in public mood are made public (to those with the ability to analyze it) faster than the effect of those changes on the market.
I'd like to know how long changes in twitter-temperament take to become reflected in market prices, as well as if these predictions have only been valuable for specific sectors of the market (I would predict consumer technology as the most accurately predicted sector, as a result of the type of people who I think use Twitter). It would also be cool to see what other chartable data the Twitter-mood can be correlated to.
But what happens when their analysis machines can't detect sarcasm in Twitter posts?
ReplyDeleteJoke aside, I honestly think that while the method is pretty dang creative, I think that the one flaw is that Twitter is hardly a representative of the public sentiment. Basic facts of statistics say that people are generally more inclined to vocalize their opinions when they are negative ones rather than positive. And there's the fact that Twitter users are not a proper sample of the general public and leave a substantial number of groups out of the sampling for the hedge fund company. Their sentiments are not necessarily represented by the Twitter population.
Maybe I'm just too cynical having taken AP Stats last year. Can this method still provide profit? Probably. Twitter may not be a perfect-world sample but it certainly represents a large span of public opinion. When the market fluctuates heavily and opinions are the strongest, that's when I expect the hedge fund company to make the most money. It's the subtle market moves that really matter, and I have no idea how well Twitter can represent sentiments during those times.
Your post was longer than 140 characters.
ReplyDeleteJoke aside, I agree with what you're saying for the most part. However, you don't acknowledge the cases in which this would be useful. The twitter userbase is probably an excellent predictor of tech stocks. Companies like Apple, and Microsoft probably get a large amount of business from the kind of people who tweet. So, if Apple releases a cool new product, Twitter will be all a-buzz with positive discussion, and this hedge fund's software would successfully detect an upswing in mood, and they would invest accordingly.
If they made their investment decisions in the tech sector, which would make the most sense, then I see a large potential for profit.
It is too bad they cannot send their stock requests by twitter to the broker so others could intercept them.
ReplyDeleteJoke aside, the hedge fund would only work in certain markets where most consumers have access to twitter and use it regularly. If they choose to concentrate on this segment for their research, they will have much better luck than trying to monitor all stocks with this. They probably also have something that keeps track of how many hits a particular subject has in a certain frequency to make sure it is not basing decisions on one or two tweets. If the hedge fund follows these guidelines and makes smart decisions with more basis than just tweets, then it will probably be successful in trading.