Now that’s ROI! How Twitter is being used to predict movements in stock markets

The case for social media ROI has been in contention ever since companies started using social media platforms for marketing, customer service and promotions. But if you’re looking for an exceptional case study of how social media, particularly Twitter, can drive ROI then look no further than Derwent Capital. They’re a hedge fund that has been using Twitter to predict movements in the stock market, by analysis of emotional words in tweets. Their method includes scanning 10% of tweets at random and then categorising these according to positive or negative moods, such as ‘alert’ , ‘vital’ or ‘happy’. This intelligence is then used to predict movements in the Dow Jones Industrial Average. Incredibly, it’s claimed their method is about 88% accurate for reflecting stock market movements, and importantly it is paying off.

The numbers speak for themselves

This method, it seems, is far from just being a quirky experiment or a novel use of a new technology. When Derwent Capital announced they were the first to launch a Twitter-based hedge fund, it seemed like little more than a PR stunt and the real business case behind it seemed questionable. Yet in their first full month of trading, Derwent Capital made their investors a 1.85% return while competitors such as S&P 500 fell by 2.2% in the same month. It seems that by accessing the emotions of millions on Twitter they’re able to beat the market and accurately predict movements which is driving a real monetary return. Now of course, it is very early days for the fund and they have not issued a statement that shows that their use of Twitter is the exclusive contributor to their positive increase. Nonetheless it is a unique case study in how social media can drive ROI, looking to the information you can access publicly, as opposed to a direct sales effort within Twitter.

This in itself is an important and invaluable example of the implication of millions of pieces of information being made available publicly. The information shared within Twitter is not siloed, but it is representative of society and there are goldmines of insights and analysis that are there waiting, to affect a range of industries and practices. What a strategy like this shows, is how conversation and emotions can potentially be directly monetised. This is often the most difficult thing to interpret when it comes to measuring the success of social media, but the benefit of this approach is that it’s using these human emotions to influence and drive the strategy, instead of looking at the effect that these human emotions have on a campaign that has been implemented to drive revenue.

Doing what Twitter can’t?

No-one has struggled more to show the monetary value of Twitter, than Twitter themselves. What Derwent Capital are doing here, potentially, is proving a business case for Twitter that Twitter themselves aren’t able to. They are showing the outcome of smart analysis of the thoughts, feelings, links and conversations shared within the Twitter platform and this is largely unique to Twitter itself. They may not be the biggest social network, but they are certainly the most public, as the majority of information in there is available to pretty much anyone, unlike Facebook where it’s unusual to come across a profile that is set to public, with personal updates ready to be searched and analysed.

It certainly may not have been the way Twitter envisaged the service going, but I think it’s time for Twitter to look beyond more conventional advertising efforts and into more unique and inherently social avenues that can make them money in ways not yet being explored. Let’s face it, every ad strategy that Twitter has explored so far extends little beyond placing a banner ad on the site that happens to look like a tweet instead of an ad. What if Twitter stop trying to place ads completely and instead look at feeding the information contained within the network – outward. The commercialisation of their firehose has likely been their most lucrative monetisation strategy to date and it shows that the value in Twitter is the conversations that you can access, as opposed to the conversations that you are trying to be part of.

The link between emotions and commerce

Leaving the numerical results of this particular example behind, the methodology employed displays a much wider implication for society overall. It shows that as a result of constantly sharing our thoughts and emotions, we are able to access insights about society that was never possible before. Sure, we could measure any element of societies that we wanted such as how much we spent on cars, how many kids we had, what age groups had certain jobs – pretty much anything – except emotions. But now, however, we’re choosing to share this kind of information with anyone that is willing to listen, and the potential this has for movements and patterns in societies and human connections to be analysed, is phenomenal. We can track public sentiment alongside wider trends, essentially taking away the element of surprise and the organic nature of human connections and emotions.

What this particular experiment shows us is that human sentiment extends way beyond the emotional and can be organised, categorised and analysed to the nth degree to plot wider trends across a range of industries. For some this may present a pretty stark vision of the future, where everything, down to individual emotions becomes trackable and measurable. But the magic that lies behind this insight is, for me, pretty exciting.