Do you know what the Google search engine was originally called? Until recently, this particular nugget of trivia had rather passed us by, here on The Value Perspective, but Larry Page and Sergey Brin initially named their creation BackRub. Clearly the world would have been a quirkier place if, whenever you wanted to find out more about anyone, you had to give them a quick BackRub.
Anyway, two decades on, this article is firmly about Googling and, specifically, a recent paper by four Norwegian academics investigating whether data from Google Trends could be used to forecast stock returns. Google searches and stock returns analysed data over the period from 2008 to 2013 with a view to finding any correlations between the two data series – and indeed one was found.
Admittedly it is a negative correlation – that is to say, the more Google searches there are on a particular stock, the poorer its subsequent performance (and vice-versa). It is statistically significant – albeit weak – but before you grow too excited it turns out that, if you devise a trading strategy around the correlation, you lose most of your returns to trading costs.
Even so, the experiment is interesting for a couple of reasons – the first being that, as a general rule, trading strategies do tend to see their returns eaten away by trading costs quite quickly. That goes a long way to explaining why, here on The Value Perspective, we have never been in the business of trying to unearth The Next Big Trading Strategy.
More importantly, though, there is a contrarian angle here. We have always been supremely indifferent to the latest ‘hot’ company – after all, the more interest people are showing in a stock, the more likely it is to be overvalued. It is among the less researched, more out-of-favour businesses that we are more likely to find value – the unGoogled, the un-BackRubbed. We trust we have got our massage, sorry, message across loud and clear.