US equities were up ~1% from last week’s close:
and closed just outside the upper 2 SD boundary of the longer term downtrend channel. However, price is at the 50% retracement level of the peak-trough range of this channel that has provided significant support and resistance in the past. We really need to take out the February highs at ~4200 before we can say that we may have a change to a bullish trend.
Despite this short-term strength, US equities were out-performed by Developed Market equities:
Long-term US Treasuries were the poorest performing asset class.
Current holdings in the Rutherford Portfolio look like this:
where we see that Tranche 4 (the focus of this week’s review) is still holding 100% Cash and so has missed out on the recent bounce (review date luck).
Performance of the portfolio since moving to the rotation model looks like this:
so we are in catch-up mode, albeit with low volatility (risk).
Let’s take a look at this week’s rotation graphs:
where we see Gold (GLD) still looking strong in the preferred top right quadrant.
Checking the recommendations:
confirms GLD as the top ranked ETF joined by DBC (Commodities – that also showed strength in the past week), VWO (Emerging Market equities) and AOR (our benchmark equity/bond fund).
Accordingly I will use the available Cash to allocate to Tranche 4 as shown below:
It is interesting to note that VWO and DBC are recommended Buys rather than VEA. As can be seen from the above rotation chart, VEA has stronger long-term momentum than VWO or DBC (further to the right on the horizontal axis) but the latter 2 ETFs are showing stronger short-term strength (more positive on the vertical axis). This is one reason that I am not convinced that I’ve got the logic quite right for the algorithm that is generating the recommendations. However, in a way, it is similar to Lowell’s BPI model in that it is selecting assets that have been out of favor and are now showing short-term strength – i.e. may be mean-reverting. I am reluctant to play with the algorithm too much/often because it gets me into the dangerous area of data snooping/fitting when I am looking for robustness in the long term – so I need to be a little patient and let it run for a while.