Backtesting an ETF Momentum Strategy (DBA, EEM, RWX, IEF, PCY, TLT)

A reader recently asked how eliminating the volatility factor from my ETF Relative Strength portfolio would change the results.  Last months results had bonds listed near the top so I reran the screen weighting 6 and 3 month returns at 50% each. The results are as follows:

First, when back-testing back to 2007, the 3 month return weighted at 40%, 20 day return weighted at 30%, and 20 day volatility weighted at 30%, the portfolio of buying the top 3 ETFs returned 86.3% since 2007 with volatility of 15.8%.  Rankings from 1-24 can be found for free on my blog.  (NYSE:SPY) returned -14.1% with 28.5% volatility.

When running the same portfolio since 2007 and weighting 3 month return at 50% and 20 day return at 50% (and thus volatility with a 0% weighting), the portfolio of buying the top 3 ETFs returned 82.9% with 21.3% volatility.  The top 3 ETFs at the end of July using this strategy were (NYSE:DBA), (NYSE:EEM), and (NYSE:RWX).

Testing the same portfolio since 2007 but changing the timeframe to 6 month return with a 40% weighting, 3 month return with a 30% weighting, and 3 month volatility with a 30% weighting the portfolio returned 51.3% with 15.6% volatility.  Rankings from 1-24 can be found on my blog.

When running the same portfolio since 2007 and weighting 6 month return at 50% and 3 month return at 50% (and thus volatility with a 0% weighting), the portfolio returned 73.3% with 20.1% volatility. The top 3 ETFs at the end of July using this strategy were (NYSE:IEF), (NYSE:PCY), and (NYSE:TLT).

Thus, in both cases by eliminating the volatility factor volatility increase in both cases.

As an aside, I am adding one ETF to the portfolio, stay tuned for my month end update to see the new ETF addition and don’t forget you can track the portfolio for free on the right side of my blog.

Written By Scott’s Investments

Leave a Reply

Your email address will not be published. Required fields are marked *