26th April 2013
This is sure to cause some excitement among those looking for a new way to crack the secrets of the stock market. A academic paper published on Nature.com suggests that using Google Trends to analyse search terms can help predict trading behaviour and could allow investors to devise a strategy to land big returns.
Tobias Preis, Associate Professor of Behavioural Science and Finance at Warwick Business School, Doctor Suzy Moat, of University College London, and Doctor Gene Stanley, of Boston University, found that using these changes in search volume as the basis of a trading strategy investing in the Dow Jones Industrial Average Index could have led to substantial profits.
The terms used among the 98 included ‘revenue’, ‘unemployment’, ‘credit’ and ‘nasdaq’ – in Google searches from 2004 to 2011.
If the search volume that day was high compared with the previous week, the Dow Jones was sold at the closing price the following day, and then repurchased at the end of the first day of trading in the week after. If the search volume on Sunday was low compared with the previous week, the researchers “bought” the following day.
By using the keyword “debt” – the term that saw the greatest fluctuation – the strategy brought in a dramatic profit of 326pc over seven years. By comparison, a strategy of buy-and-hold – purchasing in 2004 and selling in 2011 – would have yielded only 16pc profit, equal to the rise in the index during this time.
The report says: “Our empirical results so far are consistent with a two part hypothesis: namely that key increases in the price of the Dow Jones Industrials Average were preceded by a decrease in search volume for certain financially related terms, and conversely, that key decreases in the price of the Dow Jones Industrials Average were preceded by an increase in search volume for certain financially related terms. However, our trading strategy can be decomposed into two strategy components: one in which a decrease in search volume prompts us to buy (or take a long position) and one in which an increase in search volume prompts us to sell (or take a short position).”
One reason for success may well be that investors, before they sell out at a loss, search for a lot of information first.
Five observations/questions from Mindful Money
1) First this works in the US for two reasons – it is a highly networked market with a large number of stock market traders including many members of the public. Their behaviour is easily tracked by Google in a way that institutions’ behaviour or traders’ behaviour elsewhere around the world may not be. If would be fascinating to see how a similar study might work for the FTSE All Share or FTSE 100.
2) This is an historic survey from 2004 to 2011. Past studies of performance are not necessarily a guide to the future though actually we can see how this might work. Increasingly measuring sentiment in this way will play a role.
3) We wonder if professional investors and fund managers will start to incorporate this information into their thinking. Is it only Google trends or can other sources of big data have a bearing? Many investors already incorporate consumer sentiment into their investment decisions. But does the advantage of such a system erode if everyone starts to get savvy to it as a predictor of market behaviour?
4) We wonder if it would be possible to devise a stock specific early warning system? Apple anyone?
5) What will be interesting to see if anyone adopts this technique either in the UK or the US and if it works. If you decide to start doing it, remember almost all the usual rules of investing apply although with a slightly different source of information.