EquitiesSep 21 2015

How Twitter trends aid canny investors

twitter-iconfacebook-iconlinkedin-iconmail-iconprint-icon
Search supported by

Nowadays, everything happens first on social media, whether it’s breaking news, announcements from a business, or simply a provocative tweet from a media-hungry celebrity.

But when it comes to financial institutions using this platform to help guide investment decisions, surely this isn’t a viable way to dictate which stocks to buy or sell?

Well, according to the European Central Bank, maybe it is.

Its recent report, ‘Quantifying the Effects of Online Bullishness on International Financial Markets’, concluded that Twitter sentiment has a statistically and economically significant predictive value in respect of share prices in the US, the UK and Canada.

To pick a choice quote: “Daily Twitter bullishness is indeed found to be a useful investor sentiment indicator.” On top of that, 30 per cent of institutional investors admit that social media has directly influenced one of their investment decisions.

If the ECB has as good as verified social media as a credible source for investors to monitor, perhaps it’s high time we took notice. Social media is available to everyone, so we should all be keeping our eyes on the newsfeeds if we want to take advantage of the vital chatter that is evidently out there.

But when it comes down to it, how can social media help steer decision-making on the markets?

Take the Smiler roller-coaster crash at Alton Towers on June 2 as an example. As soon as it occurred, theme park visitors immediately took to social media to tell the world what had happened. At 2pm that day, updates were shared and retweeted thousands of times across different social media platforms.

It wasn’t until a whole hour later, at 3pm, that mainstream news organisations such as the BBC began to report the story.

When that happened, shares in Merlin Entertainment dropped significantly from 458.8p to 439p. As tragic as the incident was, if an investor in Merlin had been able to pick up on the spike in conversation around Alton Towers, they would have seen it as a flashing light: armed with this information, they could have changed or protected their position on that stock.

The same goes for Tesco and its announcement that it was banning sugary drinks – including Ribena – in an attempt to tackle obesity. Tesco saw a dramatic drop in its shares, losing 8p on a share (from just over 220p) within the first day. Naturally, the story had developed more and more on social media, and any investor not keeping an active eye on the news would have spotted opinion riding high on these platforms.

Data systems are emerging that can help investors make sense of the waves of chatter around companies on social media, analysing billions of online conversations every day. They can harness this and understand what people are saying about a company – both the level of engagement and the sentiment towards it.

For investors who may struggle to access highly expensive ‘listening’ systems or take advantage of a bigger investment firm’s resources, this could be priceless. A few minutes is a lifetime when it comes to making trading decisions, and failing to act on developing noise can leave people falling behind.

Looking back on these past few weeks and ‘Black Monday’, European stocks plummeted and US equity markets were bracing themselves for the worst after Chinese shares experienced their worst day since 2007.

When it comes to macro factors like this, the world is watching and panic ensues when investors all try to make their moves at the same time.

Of course, this was sparked by the Chinese central bank devaluing the yuan, and it was hard to miss. The impact of this as we move forward, however, is that there is obvious momentum and people are keeping their ears to the ground far more than usual when it comes to the stocks they hold. Those who want to stay one step ahead would do well to look beyond the obvious channels.

If the ECB has woken up to the benefits of social media listening, and so much so that it explored it in the level of detail its report offered, you can bet your bottom dollar it is already being used beyond retail investors.

Technology moves quickly, so don’t be surprised if big institutions, pension providers and fund managers alike are using these innovative tools to supplement their traditional monitoring systems. With social media so ingrained in our daily lives, there’s no reason we shouldn’t all be making the most of it in this way.

Gareth Mann is chief executive of Trading.co.uk

Online sentiment: The ECB’s view

In its report, ‘Quantifying the Effects of Online Bullishness on International Financial Markets’, the European Central Bank noted that daily Twitter bullishness can be a useful indicator of sentiment.

The report concluded: “The reliability and accuracy of existing computational measures of investor sentiment leaves much to be desired. We therefore propose a direct and unambiguous measure of investor sentiment, namely the relative frequency of occurrence of two terms commonly used by investors in Twitter updates and Google search queries.

“Our analysis shows a positive correlation between Twitter bullishness and Google bullishness on a weekly basis; furthermore, it finds that the former leads changes in the latter... More importantly, we find that daily Twitter bullishness leads stock index returns in the US (Dow Jones, S&P 500, Russell 1000, Russell 2000), the UK (FTSE 100) and Canada (GSPTSE), but has only very modest predictive value in respect of the Chinese stockmarket.”

However, the report acknowledged that while high Twitter bullishness “predicts an increase in stock returns, we observe that these return to fundamental values within a week. Our research thus appears to support the hypothesised role of ‘investor sentiment’ in behavioural finance.”

Definition: Twitter bullishness

The ECB states in its report: “We define a tweet as bullish if it contains the term ‘bullish’ and bearish if it contains the term ‘bearish’. Over the study period from 2010 to 2012, we find about 0.31 million bullish and bearish tweets. There are 1,091 days in total, and the average daily number of bullish and bearish tweets is 280.”