Crocodilian

First Call of a Double-Dip Recession: Setting Up a Market Bottom?

Very creative use of Google to measure zeitgeist, but this mode of analysis has the potential for rapid feedback distortions. That is: Have you considered that your contrarian schema may itself propagate? The "velocity of opinion" has increased dramatically in the Twitter age.

Also, do you have any idea what accounts for the dramatic periodicity in your Google data? Your chart shows a huge annual cycle, with troughs in the summer months. Is this raw search volume? If so, you should really normalize against overall search volumes (eg, what percentage of searches are for "Great Depression", as opposed to the nominal search numbers).