How Google Trends and Internet Searches Correlate with Asset Prices

How Google Trends and Internet Searches Correlate with Asset Prices

We’ve taken a fun exploratory look at internet search terms and their potential correlation (or lack thereof) with financial asset prices before (See: Using Google Trends as a Leading Indicator).

That piece actually found its way to Google Headquarters in Mountain View, as I was contacted by someone from Google Finance who asked if I’d done any additional research on the topic!  And I unfortunately botched that connection by not figuring out how to access my email address for this site until two months later!

It’s a topic I’d like to continue to explore, so I was very interested to see this guest piece on The Big Picture entitled Google Trends – A fresh look at search frequency.

“Stock market crash” was a leading indicator.

An interesting case study of this is a search for the term “market crash” or “stock market crash”, backed up to the very start of the market crash itself.

This is the graph of search frequency for “stock market crash”, which is largely similar to that for “market crash” (chart):

Note the clear peak in September 2008, with nothing really before that. The January 2008 spike was largely the same as what we saw in ~February 2007.

The simple response to this is “oh yeah, this simply peaked while the crash was happening”… but that might be a false assumption.

Digging into the weekly data seems to paint a different picture, interestingly enough. We got a reading of 4 in the chart above on … September 15th 2008 (chart):

On September 14th, the S&P was at 1255. Heck, even when Search Terms went to 6.4, the S&P was still at a fairly rich 1099.

In other words, a spike to 4 actually *led* the crash, and was *not* a lagging indicator after the fact.

The author also explores the recent housing bubble and its search relationship, while looking ahead at what current hot terms “double dip” and “bond bubble” may have in store for the future.  You can read the full piece here.

  • heyy Brett..I was doing some research on this myself and although I haven’t formally tested this hypothesis..but i think its plausible that Google data can be used to predict market that I mean that you can’t predict the exact number..but you can readily tell whether the average return in the coming week will surpass the previous one..ive tried it for a few companies (geoeye and ediets)…the trick to getting good results is building a proper model rather than just relying on search nowhere near hacking it yet though….im just waiting for my CFA exam to be over so i can go over this properly..if you  are interested contact me @…again…good read!

  • brettowens

    Thanks! I sent over an email, hopefully it went through. If not you can reach me at

About Author


Hi, I’m Brett Owens – and I’m a financial junkie. My “problem” started incollege, when I got a little dose of the stock market – man, was I hooked…in no time, I was reading the Wall Street Journal religously.

Sign up for our Newsletter