Can search volume be used as a variable to predict share prices?
I recently came up with an idea to analyze search volumes and compare these to the increase/decrease of listed companies on the JSE, to determine if there is any correlation.
When you consider search volumes it’s an indication of Interest in a company by all stake holders: Investors, employees, customers and the general public. While the share price is an indication of the market environment as well as investors confidence in the company.
The analysis is time consuming and very manual, therefore i have just limited the analysis to a single company (MTN) for the last year.
Search volume data was extracted from Google insights tool, while the past years share price history was extracted from Standardbanks securities website. Since Google provides share volumes summarized per a week, a weekly average of the daily closing price was used.
No clear patterns emerge from the graph, however there is an increase in search volumes between week 5 and 10 (April 2008), we also see the share price increased to it’s highest point in the past 12 months. This was around the time there were talks of MTN being taken over. Though further down, there seems to be a negative correlation. The correlation coefficient is actually -0.522484149, i.e the correlation is actually negative. This is counter-intuitive. Though you have to take into account the current financial system and most stocks have lost +-50% of their value since last year this time.
When i get more time, i will look at more companies however I think further detailed analysis with search volumes vs share prices could uncover some relations, what are your thoughts?
Disclosure & stuff for the lawyers
I currently own shares in MTN, and this does not constitute financial advice.
If you would like to take a crack at the data, you can find MTN share volumes vs share price data on Google docs.