This is a valid question because there are so many articles about how social media, and Twitter in particular, are able to anticipate movement on the stock markets. With 320 million users (as at 6 May 2016) and 500 million tweets sent per day, Twitter has the advantage of being both instantaneous and volatile. For some, it has become the go-to place for share price analysis and recommendations. Several studies appear to support this theory, the best known of which was conducted by IT researchers at the University of California, Riverside, together with a Spanish team. During the four months from March to June 2010, they analysed the activity of 150 companies chosen at random from the S&P 500 Index, and noticed a correlation between the number of tweets they sent and their share price. The more positive and greater in number the tweets relating to a company, its CEO or its products, the more that company’s share price would rise. Having observed this, the researchers devised a mathematical model, applied it to an imaginary portfolio and outperformed other financial strategies based on financial analysis by as much as 11%.

It only takes one tweet…

Does this mean a single tweet can prompt share prices to crash or soar? It would appear so, at least in the US, where firms spend close to €1 billion on controlling their Twitter reputation.  In September 2015, Hillary Clinton, then in the midst of her US presidential campaign, reacted strongly on Twitter following a biotechnology company’s decision to increase the price of a drug treating toxoplasmosis and malaria by more than 5,000%. In her tweet, the presidential candidate merely promised to crack down on the pharmaceuticals market. Just minutes after her 140 characters were published, the Nasdaq biotech index had dropped by 4.5%.

There are plenty examples of this type, to the point where analysts in the US have specialised in the spreading of tweets to change share prices and make money. Closer to home, in Luxembourg, a financial engineering start-up is developing financial prediction algorithms based largely on analysing social media (not just Twitter). This would make it possible to determine the direction and scale of stock market movements in the hours following analysis of sentiment expressed on social media.

Not always a reliable indicator

In future, will you just need to look at social media to invest successfully on the stock market? Absolutely not. Social media may eventually become a useful analysis tool for finance professionals, but they are by no means infallible. The main reason for this is that data on social media are not always reliable and relevant. They contain myriad unfounded rumours and lies. These might momentarily disrupt a share price, but it will stabilise after a few days, or even hours. Moreover, given the speed at which information on them is spread and interpreted, social media clearly should not be used by individuals, even if they are informed, but rather by traders using highly speculative strategies and seeking short- or very short-term returns.

Last, and most importantly, social networks are not permanent entities. Twitter might even cease to exist sooner than people think. The company is currently going through a bad spell, losing money for its shareholders and those who want to sell it. Discussions are under way with many large groups, including Google, Microsoft and Disney. There is no guarantee that these talks will lead to a sale. Equally, there is no guarantee that Twitter will operate in the same way under new ownership.

Even the algorithms themselves, regardless of whether they are used in social media, can have weaknesses. A recent study by the University of Cambridge showed that professional traders were able to outperform the most sophisticated algorithms simply through gut feelings!


We use cookies to enhance your experience. Basic cookies are essential for the proper working of this website. For example, they save your language preferences. They also help us gather anonymous information about the use of our site. More information in our Cookie Policy

I don't agree I agree