- Alexander HornKeymaster12/28/2014 at 1:34 pmPost count: 383
Yes, we have replicated and backtested Gary’s great work, btw a must read. But here we feel rather: ‘Laurels to whom deserves them’, and we’re not planning to include plain replicated strategies into our subscription service if we cannot provide an added value to them. Hope this sounds fair and rational to you.
If you are interested to include it into your portfolio you can find a great and straight forward implementation in a google sheet at Scott’s Investment.
- Alexander HornKeymaster01/08/2015 at 1:21 pmPost count: 383
I’ve just looked at the model I made back in 20113 based on the ‘Risk Premia Harvesting Through Dual Momentum’ paper, and came to 13.8% CAGR for the 1974-2011 period used in the paper. This is quite close to the published 14.9%, so my ego was pleased.
But I must say that I used the MSCI data and other index data mentioned in the paper with some simplifications, e.g. this is not really an investable version. I’ve not tried with ETF, as at the end I moved on to other promising ideas.
- sfreewellParticipant01/08/2015 at 9:15 pmPost count: 3
I’d be curious to see what your results are using etfs. I have a sneaking
suspicion that results using etfs will be far inferior to using index data.
I realize that you can’t test back to 1974, but you need to test with tradeable
- Alexander HornKeymaster01/08/2015 at 11:21 pmPost count: 383
Yes, most probably the results are less (because of fees) and different (because of tracking errors, index used). I’d love to give the model an update, but honestly we have some nice other strategies in development (spoiler here), so I rather add value there.
Have just looked for recent backtests after Gary’s new book came out, and maybe this is a nice one:
And some other articles: http://itawealth.com/?s=dual+momentum
And as mentioned above, ‘Scott’s Investments’ has also done some backtests.. http://www.scottsinvestments.com/dual-etf-momentum/
If not, here the ultimative link (with a smile): http://bit.ly/1wAr1b1
You must be logged in to reply to this topic.