Why a reset?
Today, on the first trading day of march, I’ve decided to reset the Trend and Swing Portfolios because in the past few months I have made significant changes to the algorithms and I want the back-test results to match the new changes. Here is a brief list of on going changes that were made since November 2013:
- All Portfolios now behave better in out of sample periods which should make them more consistent in the future.
- Trend Portfolios now buy on a price pullback like the Swing Portfolios but they stay long until the trailing stop is hit.
- Both Swing and Trend Portfolios now check for correlation before buying another stock to improve diversification. Note that the parameters are a bit lax so it is still possible to have a many stocks in one sector.
- Both Swing and Trend Portfolios now have a volatility criteria. If a stock is not volatile enough or too volatile, it will be ignored.
Here is a recap of the first “live” 6 months performance of the Swing and Trend portfolios, I am very happy with the results even if I fully realize we are in a big bull market (up to February 2014 at least!) and there were a few bad trades in the CAD Swing portfolio.
Live Performance from August 23rd 2013 to Febuary 28th 2014:
- US Trend: 58.4%
- CAD Trend: 30%
- US Swing: 25.4%
- CAD Swing: -4.6%
For the same Period SP 500 returned about 11%.
I have created an archive to preserve that time period here.
I’m not sure how the stock markets will behave for the rest of 2014 but I think it will be more volatile and less bullish, but I hope the improvements done to my systems will compensate for it. For example, the annual growth rate of the US Trend Portfolio jumped from 18.9% to more than 27% while reducing the draw down to 20% from 22% for the past 10 years period.
Note that many positions tracked by the Portfolios last week will have disappeared, this is expected since the time line of all portfolios will be reset and buying rules were changed.