First post on QC since migrating from Quantopian and wanted to push out to the community a strategy framework we have been using on the bond side and multi-asset side for retail investors in separately managed accounts live since 2010 .   This is not going to have a massive return profile but is used as a balancing mechanism.  I typically like the barbell approach for risk where we have a high risk reward on one side and lower on the other,   If you haven't read Antiragile, Things That Gain from Disorder by Taleb, I highly recommend it.  

With this in mind, and the fact that we are potentially facing an era where bonds may underperform compared to the last 10 years and I thought the community could benefit from an approach that dynamically selects one or more bond ETFs.  The ETFs cover the full range of yield, quality and maturity and will adjust based on three factors based on the rebalance timeframe.  

  • Short term return timeframe
  • Longer term return timeframe 
  • Volatility.

 

These are weighted and ranked with volatility being a governor….the more price volatility in a specified time frame along with the weighting will lower the security in the rankings.   Note the moving average is not a typical moving average and is based on the closing price of the last day-last N months.  This is very old school simple approach (1920's era), and removes a lot of signal noise but is also and area for more research. 

You can run this framework within single asset classes, multiple asset classes, add risk management, dynamic universe selection or leverage.  Make it your own.   

A special shoutout to Jovad U. who assisted me in bringing this to QC.  

Author