UsableMarkets

Not sure how I didn’t see this before, but Google Finance now has a new(ish) feature called Google Domestic Trends. This is essentially Google Trends, but with the searched for keywords organized into useful concepts (i.e. retail, durable goods, automotive, etc.) and hooked into Google’s stock charts. This way a user can take a concept like travel and chart search frequency of travel related terms to a stock like Royal Caribbean Cruises (NYSE:RCL).

Percent change in RCL stock price and number of times Google search users searched on travel related terms, 2006 to present

On their own, without mapping to a stock, some of the trends are interesting, if perhaps expected. For example, searches regarding unemployment have dramatically increased over the past two years.

Change in number of times Google search users searched on unemployment related terms, 2004 to present

While exploring search terms in a conceptual manner makes tons of sense, and can actually say something interesting (even if it conforms what we already knew, like people’s interest in real estate falls off in the late fall and winter, and picks up in the spring) …

Change in number of times Google search users searched on real estate related terms, 2004 to present

… it’s a bit of a stretch to compare these to the stock price of a specific company. For example, TDAmeritrade’s stock price does not appear to correlate with the number of times people search on investing related search terms.

Percent change in AMTD (TDAmeritrade) stock price and number of times Google search users searched on investing related terms, 2006 to present

This holds true with multiple concept vs. multiple stocks (although I won’t claim to have made a scientific study out of this).

While Google is perfectly within their rights to explore how their different tools and data can work together, this mashup seems a bit pointless. While in theory search mindshare might ultimately be reflected in a specific company’s stock price, it’s a bit of a stretch to find a correlation between the two, much less an underlying factor causing the two to consistently move in tandem.

But, one might argue, what about mapping these trending concepts to economic indicators? For example, mapping unemployment searches to the actual unemployment rate. This doesn’t seem possible in the current interface, but if we download the data to excel and compare it to the unemployment rate produced by the Bureau of Labor Statistics, we can see the trends map closely, with the Google unemployment search trends exhibiting more volatility than the BLS rate.

Percent change in unemployment rate (from the Bureau of Labor Statistics) and the number of times Google search users searched on unemployment related terms, 2005 to present. click for larger size

So, overall, a useful tool? Undoubtedly, and seemingly appropriate at a macro level.

Comparing to stock prices? Mmm. Not so much.

~alex