Historical Drawdowns as a Roadmap
If the market 'experts' want to run through red lights we might as may look for some kind of a road map to let us know what intersections are the most dangerous. We can draw a map by putting this year's rally from the March 9th lows in a historical context to help evaluate potential future movements in equity and volatility. Using data since 1928 from the Dow Jones Industrial Average I ranked the 10 largest maximum drawdowns in history. A maximum drawdown is an extremely useful measure of risk and is defined as the greatest loss from the highest peak to the lowest low over the history of any asset, fund, or index. I prefer the concept of a maximum drawdown in evaluating quantitative strategies with the goal of generating the highest projected annualized return per unit of expected maximum drawdown. The annualized return/max drawdown ratio, alternatively known as the MAR ratio or Calmar ratio, provides a means to measure and compare the effectiveness of different strategies or funds.
After generating the list of the top 10 largest drawdowns in history, I looked at the rebound in the market following each drawdown and chronicled the length and magnitude to its highest point subsequent to a 10% correction. From there I evaluated the performance of the DJIA six months from the peak of the post-drawdown rally. Over each market phase I also measured the rise and fall of the 21-day statistical volatility of the DJIA. The results of the analysis are presented in the table below. This includes an analysis of the distribution of drawdown magnitudes and changes in statistical volatility compared to the logarithmic changes in the DJIA index.
The drawdown experienced from 2007 to 2009 is one of the most extreme on record, second only in magnitude to the 1929 crash that initiated Great Depression. It is worth noting that it is possible (but unlikely) that the current drawdown is not even finished, and that we are in the midst of an extended bear market rally and may eventually see the March 9th low taken out sometime in 2010 or 2011. Keep in mind that during the Great Depression the three year drawdown period between 1929 and 1932 involved at least six bear market rallies (according to my count), some lasting as long as six months. Although extremely unlikely given the amount fiscal and monetary stimulus in the economy it never hurts to prepare for the possibility that an extended multi-year decline could extend the current drawdown.
Taken in the context of the largest historical drawdowns, the 50%+ rally we've seen over the past 6 months and subsequent 70% drop is statistical volatility of the DJIA is not at all unusual. The average post-drawdown rally lasts 14 months with an average return of 54% before a 10% correction. On average during this period statistical volatility fell 46%. The performance of the market six months after the post-drawdown rally is decidedly poor, with an average return of -7.38%. In almost all instances, the market moved sideways for months following the high of the post-drawdown rally before starting a new bull or bear market move. The six months following the post-drawdown rally statistical volatility averaged an approximate 27% premium over the measure observed at the height of the rally. This would seem to indicate that current buyers of portfolio insurance may be justified in demanding elevated premiums over statistical volatility in the back months of the volatility term structure. We have indeed reached an important and perhaps increasingly dangerous intersection.