Instead of combining all the review material for the Einstein into one blog post, I’m breaking this update into two parts. The first part will focus on cluster analysis and the second part will explain how I interpret the Momentum-Optimization Model.
Cluster Outline: To fit the clusters into ten groups, the cutoff was set to 0.78 instead of the customary 0.80.
Einstein Rankings: Note I am using the semi-variance calculation, one of several options. The choice of SV_60 does alter the ranking among the top ten ETFs, but does not work its way down the chain to the lower performing ETFs. U.S. Equities continue to hold most of the top spots.
Cluster Buy-Hold-Sell Recommendations: If one is following the cluster model, a significant percentage of the portfolio is concentrated in VBK, IDV, and JNK. These ETFs also showed up when the Aristotle was reviewed yesterday. If you look at the Cluster Outline in the first screen shot you will see that VBK is one among many ETFs. Personally, I’m more comfortable “spreading the wealth” into other ETFs, particularly some of the dividend oriented ETFs. When I run the Momentum-Optimization Model in Part II, readers will see how this works.
The argument in favor of Clustering is that diversity comes from linking low correlated securities, not spreading the portfolio over a wide range of investments. Finding low correlated ETFs is where the search HedgeHunter is conducting will be useful.