Gavekal Capital: Stocks Have Gone on Another Statistical Run

Thursday, June 26, 2014

Stocks Have Gone on Another Statistical Run

Many of our readers are probably familiar with the concept of a "random walk" in which stocks are just as likely to go up as down on any given day (i.e. the outcome is a coin flip). Empirical analysis shows that this "random walk" assumption generally holds over long periods of time, but that over short periods of time stock market outcomes can deviate meaningfully from 50:50.

To help illustrate this point, think of a baseball player who over the course of a season bats .300, but occasionally goes on a month-long run batting .400 or a month-long slump batting .100. In a similar fashion the stock market can go on "runs" in which over the course of four months 65% of the trading days can post gains or "slumps" in which over the course of four months only 40% of the trading days can post gains.  

In the chart below we measure exactly that: the percent of days over the previous four months in which the stock market has gained. What we find is that over a four month observation period the stock market rarely posts gains more than 60% of the time (a statistical run) and when it has it is usually followed by a period in which the market only gains 45% of the time (a statistical slump).

We point this out because we are currently in the process of coming down from a statistical run that appears to have ended on June 9th. If history is a guide, we should now expect a more random outcome of stock market performance over the next several weeks and months and may even experience a period of statistical slump. Note that runs and slumps don't always coincide with massive price gains or losses, but this chart shows that it is possible.