Social Sharing
Extended Viewer
In the Market - June 2009
Jun 15, 2009
Maintaining, Sustaining, and Retaining Diversification
Following up from last month's newsletter piece we are taking a further look at equity diversification. Last month we noted that in periods of high market losses, correlations between equities within the Dow Jones rise to the point where equity diversification ceases to exist. Can a portfolio be constructed that exhibits diversification in bad markets while still producing acceptable overall returns?
We looked at the Dow components dating back to 1997 and studied how different groups of positions interacted with each other during periods of low volatility, mid level volatility, and high volatility as illustrated by the VIX. Breaking the Dow components into 4 sub groups; highly correlated with low volatility, highly correlated with high volatility, less correlated with high volatility and lastly, less correlated with low volatility, we looked at how each of these groups performed during the three periods of volatility described above. What we saw was the extension of the work performed last month. The high corr/high vol did great in low volatility (bull market) regimes but performed terribly during high volatility (bear market) regimes. The low corr/low vol portfolio performed well in bull markets but not as well as the previous portfolio, and while it performed poorly in bear markets, it did not perform as badly as the previous portfolio. This is not surprising. The logical next step was to see if the low corr/ low vol portfolio performed well enough in the bear markets to cover the difference in returns that the high corr/high vol portfolio receives in the bull markets.
To test the theory, we looked at all the components of the Dow Jones Industrial Average on a year-by-year basis. We computed the correlations between each constituent and their volatility. Comparing these numbers, we put together a list of the 5 constituents that exhibited the lowest volatility and lowest correlation to the other constituents. Those 5 stocks would be the basis of our portfolio for the following year with equal weighting. We repeated this process every year and developed a time series for our portfolio. We looked at this portfolio vs. the equally weighted portfolio of the Dow 30 during the same time periods. Below you will see 4 charts that show the relative performance of both those indices. The top left chart is for the full time period, and the other 3 are for the different volatility regimes discussed above. You can see that results matched what we were expecting; the full portfolio performed better in bull markets, but lagged our low volatility, less correlated portfolio in times of market distress. During the mid volatility regimes, the 2 portfolios seem to have similar returns but as the volatility starts to grow our portfolio starts to outperform. Finally, in the high volatility regime we can see that both indices are down but our portfolio holds the earlier gains better. Overall, our low vol/low corr portfolio out performs an equally weighted Dow portfolio.
If the investor has a desire for greater returns during the low volatility periods, leveraging the portfolio may help to increase gains seen during rallies. Over the full 11 years of the analysis our low volatility, less correlated index has a lower risk as measured by volatility and a significantly greater return than the Dow Jones.
Download