Another example is yield spreads on bonds. Bond traders look at historical yields, calculate the average, and then identify profitable trade situations where the spread is different from the average.
But "mean reversion" shouldn't just be put in an investor's toolbox as a substitute for deeper thinking. For example, another cousin of mean reversion is "doubling down." But fortunes have been lost by "doubling down." After all, if you liked a stock at $30/share, you've got to love it at $15/share. Talk to some of those who tried this during the dot-com bust of 2001!
In thinking about mean reversion, you should think through the underlying process. Essentially, we're back in high school studying concepts like stable functions, limits in Calculus, etc. Take a step further and we are on the fringe of Chaos Theory. These concepts teach us that some functions effectively blow up whereas others reach a limit. In the world of economics, some events bring forth actions that move us back towards equilibrium whereas others blow up. Further complicating the process is the possibility that the mean may change.
An example of a blowup is the 2008/early 2009 period in the U.S. stock market. At one point, there was a run on money market funds reminiscent of the bank runs on the 1930s that threatened to take the financial markets down, i.e ,to literally blow up the system. As expected, forces came into play to rein the process in - Bernanke and Paulson guaranteed money market funds.
A better example is Japan of the 1990s which morphed from the fastest growing major economy in the world to what can best be described as a basket case. To this day, it hasn't pulled out of its malaise. Maybe the mean it has reverted to is permanent weak performance and the exceptional performance of the earlier period was an aberation. This, of course, is little help to investors who pounced on Japan's stock market seeing exceptional value after it moved sharply lower.
The bottom line is that thinking along the lines of mean reversion leads to conceptually coming up with an idea of where the world should be. In effect, it is a type of linear thinking in a world that mostly isn't linear. Although useful in many places in framing investment strategy, it has to be handled with care.
One final example worth contemplating along these lines is the so-called problem with asset correlation. By diversifing among widely varying asset types, negative correlations supposedly would lessen risk in a downfall. Worked great in a non-globalized capital markets setting. But, as money moved more easily around the world, the historical correlations fell apart. Think through the underlying process!