This post is a bit different to the other material I’ve submitted on this blog. I wanted to try and evaluate different strategies for sector/asset class rotations based on certain criteria such as performance etc. I have been playing around with a few approaches. Here is an example …
Universe of underlyings restricted to select ETFs spread across the “Broad Market”, “Sectors”, “Commodities”, “Fixed Income”, “Currencies”
Filtering Criteria (Creiteria 1)
* Price > SMA 20
* 20 < ADX <= 40
ADX = Average Directional Index
Looking to capture underlyings which are trending, but not excessively deep into a trend
Sort the list generated by Criteria 1 in descending order by the following metric …
[3 month return] + 0.5 * [10 day return]
– Initiate a long position in the top 3 underlings identified after applying the 2 filtering criteria mentioned above.
(Split investment capital evenly across the 3 underlying)
– Re-assess every 2 weeks (10 business days)
– Close out any of the initiated positions falling out of the top 3 bracket and replace the associated dollar amount with ones that are appearing in the top 3
From 1-17-2008 to 1-18-2013 (5 years) – Perf 1
From 1-2-2001 to 1-18-2013 – Perf 2
Note: Not all underlying I am using have been trading since 2001. Some started fairly recently.