Can Twitter tell you how to invest?

New algorithm based on social media platform helps harness the power of hype to secure financial gains – what could possibly go wrong?

Tweets fond of individual shares have been found to influence the price (Shutterstock)

Social media. It has eased its tendrils into every part of modern life. Suddenly the global village is right there, 24/7, broadcasting, judging and commenting on everything from pandas falling over to the latest troop movements in Syria.

Increasingly, the very modern phenomenon is more than a reflection on real life – it is creating the noise, the sentiment, the ‘real’ life.

When it comes to assets, regardless of what they are, where they are or how you deal in them, value is only ever what someone is prepared to pay for them at a particular point in time.

So, while experts will always urge us to examine the underlying strengths and weaknesses of something we may or may not invest in, hype is the intoxicating smokescreen hiding that truth.

Create a conversation about something and you can affect its price. Think Bitcoin, for example.

That’s hardly news. But where are those conversations now happening? Twitter.

Joining the conversation

Follow this logic and you can start to see why one group of academics in Rotterdam have decided that Twitter posts could be used to predict the stock market, helping, they argue, traders and large scale professional investors to make better decisions.

Ting Li, Professor of Digital Business and Dr Jan van Dalen, both from RSM Rotterdam School of Management, Erasmus University recently analysed over a million Twitter messages that mentioned stocks listed on the S&P 100 share index.

The researchers then developed an algorithm that looked at the sentiment of the tweets and extracted distinct ‘buy’, ‘hold’ and ‘sell’ signals embedded in them – before comparing them to actual price fluctuations on the stocks over the following days.

They found that stocks that are tweeted bullish sentiments such as ‘buy!’ experience, on average, higher abnormal returns.

The results also showed that the relationship between bullish language and increased stock performance was even stronger for influential Twitter users who are frequently retweeted and often mentioned – and that the number of tweets about a particular stock could predict trading volumes, volatility and follow-up return on a stock.

Often the more that Twitter users disagreed about a stock, the higher the trading volumes.

To test if these findings could be the basis for a profitable trading strategy, the researchers ran a simulation using the information from the study and found that, even taking transaction costs into account, the simulated returns beat the market.

“This study shows the potential value of information on Twitter for making informed trading decisions. In the simulation we showed that if you invested money in the S&P 100 and used the information gleaned from Twitter using our algorithm, you would beat the market. You could invest in one company or a number, you could sell at the end of each day and reinvest the next or you could trade every other day or every 3, 4 or 5 days – and even when you take transaction costs into consideration, you’d still come out ahead.”

“This could be used by institutional investors or home-based day traders and proves that Twitter isn’t just noise – useful information can be extracted and could help investors make better decisions,” claims Li.

Seems plausible...

But if you were beginning to wonder whether Twitter could hold the key to all of modern life’s first world challenges, don’t put common sense on ice just yet.

“The research is interesting and I certainly don’t envy anyone for having ploughed through over a million Twitter messages in the course of their research,” says Tom Selby, Senior Analyst at AJ Bell, who notes the simulation ran for just 21 weeks.

“This feels like a bull market phenomenon to me and something which may work for a while – until it doesn’t,” he adds.

“The bulk of Twitter’s existence since launch in 2006 has covered a steaming bull market which has taken headline indices to levels way above those seen 13 years ago, so a lot of boats will naturally have risen with this tide. It will be interesting to see how things pan out when the next bear market hits and stock prices start going down again.

“This is far from a tried and tested strategy and on the face of it looks like a recipe for disaster,” warns Laith Khalaf, senior analyst at Hargreaves Lansdown.

“Some of the worst cases of investment loss have been sustained by following the herd and over-trading and this approach has both of those elements in spades. Investors would be better off sticking to a long-term investment strategy, placing their money into a diversified portfolio and trading as little as possible.

“Don’t just do something, sit there!”

But this research also confuses a crucial distinction between traders and investors, one that seems increasingly blurred.

Investors buy stakes in companies because they believe those firms have a strong competitive position, a good strategy, strong finances, competent and trustworthy management, a sensible valuation and the potential to pay consistent dividends over time. Those dividends help investors turn time into money.

“Traders are different,” says Selby. “By using sources such as Twitter for information, they are looking to judge short-term sentiment towards a stock and try to make a profit by timing the movement of money. It’s an entirely different strategy and the two should in no way be conflated. ‘Investing’ will suit some people and their financial goals, needs, time horizon and risk appetite while ‘trading’ will be more appropriate for others, depending on their personal circumstances.

“Investors – and traders for that matter – should proceed with caution if they think Twitter is a potential source of easy money because no such thing exists. Those traders (or investors) who were lulled into a false sense of security by the bull markets of 1998-2000 and 2003-07 only to be badly burned by the bear markets in 1998-2000 and 2007-09 will be able to remind you of that.”