For the convenience of those interested in tracking Mebane Faber’s popular Global Tactical Asset-Allocation model that he originally featured in his 2006 paper, I have added that model among our tactical ETF models. As I discussed in a Seeking Alpha article, I have replaced the original 10 period moving average with the 8 period…it just seems to work better. Overall though, readers will find this tracks pretty close to what Meb has on his own site.
As some of you are probably aware, Meb has published some updates to his original paper, the most recent being February 2013. In the latest update, Meb introduced 8 additional asset-classes for a total of 13. He also discussed several alternative methods of allocating funds, one which included the discussion of combining a momentum-based ranking system along with the original use of the trend following moving average. (Links to both the original and the updated paper can be found on the Reference page.)
This inspired me to conduct some research of my own given the tremendous success I have had with applying momentum across a variety of asset-classes. After testing many combinations of factors, the finished model ranks the 13 ETFs using a ranking system that combines both momentum and volatility. Any ETF that is not above the 8 period moving average is eliminated. Finally, instead of selecting the top five ETFs as Meb does in his paper, I settled on buying the top three in equal weight as the best choice. Below in the table are my test results.
|Number of ETFs||CAGR||Sharpe Ratio||Maximum Drawdown||Total Return (dividends reinvested)|
To summarize, less than three and the portfolio volatility reached unacceptable levels and more than three, the portfolio returns quickly diminish. Thus, I have aptly named the new model, GTAA – Focus Three.
For some who may think that owning just three ETFs each month is too limited, remember that each ETF is in itself extremely diverse. For example, the iShares Russell 1000 Growth ETF, a current model holding, is diversified with 625 separate securities.
I personally really like the stability of this model and how well it navigated the the 2007-2009 period. Unlike some of my other models, this model does not include an inverse fund but never the less, the inclusion of fixed-income ETFs provided enough opportunity to keep downside risk contained to acceptable levels.
For individual investors looking for a single model that covers most all of the major asset-classes, I believe this could be a good choice. And for financial advisors who want a simple but effective way to managed their client book, I would have to think this model could serve them well and provide the ability to build a very scaleable practice.