There are a number of studies which have shown relative strength to be an important factor leading to portfolio outperformance.Portfolio Cafe embraces these studies and fact, utilizes this factor in a number of the model portfolios.
Dorsey Wright & Associates has been one of the leading firms in relative strength research and they have just produced another white-paper which I have added to the Reference page.
The paper, Tactical Asset Allocation Using Relative Strength, by John Lewis, CMT, is one of the first papers to discuss the use of relative strength or “momentum” as it is sometimes referred to as, with baskets of ETFs.
For those wanting just the CliffsNotes, here you are:
Testing was done on 100 ETFs spanning a variety of asset classes and ETF sponsors. The universe composition included traditional asset classes as well as commodities, currencies, and even inverse ETFs.
Portfolios were constructed containing 10 highly ranked ETFs.
In order to determine if “luck” was involved in the selection process, numerous portfolios were created and Monte Carlo analysis was conducted to test the robustness of the selection process.This was done by randomly selecting from among the top quartile, the final 10 constituents.
Tests were ran for the period 12/31/1999 to 12/31/11
The Results
The 10 ETF portfolios outperformed the S&P 500, the 60/40 Balanced Index, and the Barclays Aggregate Bond Index 100% of the time, even when securities were drawn at random from the top quartile.
The minimum return was 121.60% versus the S&P 500 return of -14.1%.The median return was actually 197.4% which was also close to the average return of 198.80%.
Other Conclusions
Their studies showed that relative strength performed best when intermediate time periods (3 months – 12 months) were used.
Relative strength does a good job of managing volatility because of its adaptive nature.When volatile asset classes are declining, it might rotate to less volatile asset classes like bonds and currencies.
And perhaps the most important conclusion that can be drawn from this is that it is certainly possible to achieve better returns than those offered by the standard broad based benchmarks.