Occasionally, I’ll encounter an ITA reader who is interested in constructing a portfolio around sectors. Those who have read Gary Antonacci’s book, Dual Momentum Investing may recall him stating the following. “My favorite dual momentum strategy is one that rotates among the strongest U.S. stock market equity sectors.” This blog is a multi-level explanation of how one might employ Antonacci’s Dual Momentum Model using sector ETFs.
Main Menu: Back-testing has shown that the 60- and 100-Day look-back periods provide superior performance compared to the 252-Day look-back advocated in the basic Dual Momentum Model. Therefore, I am using the “default” settings from the Kipling Tranche spreadsheet. Check the percentage weights assigned to the different periods plus the volatility factor where 20% is applied. What these shorter look-back periods do is alert us to market changes considerably faster than the 252-Day look-back time frame.
Sector Tranche Momentum Recommendations: Using information from the above Main Menu, the ten (10) sector ETFs plus VTI are ranked. ETFs immediately eliminated from consideration are VDC, VNQ, VHT, and VPU as all rank below SHY when using the absolute momentum screen. The relative momentum ranking system identifies the top five in the following order: VFH, VIS, VAW, VOX, and VCR. One can stop right here and invest 20% in each of these five ETF. If five sector ETFs did not rank above SHY, we would only consider using those that did. If none ranked above SHY we either go 100% to cash or investing in SHY. This move keeps us out of deep bear markets.
However, we move on to the next step or a further refinement as to what ETFs to include for consideration in constructing the sector portfolio. Move over to the REDA Expectancy part of the table. Here we are looking for sector ETFs that have a Group ranking of 1, 2, or 3. As mentioned in prior blogs, finding a number one (1) ranked ETF is rare and we see none in this example.
A sector investor using the REDA screens would invest 33% in each of VDE, VFH, and VIS. This is our second stopping point. However, there is one more screening process, should a sector investor wish to test one more refinement. See the Point & Figure (PnF) graph.
The last test is to see if either VDE, VFH, or VIS is outperforming VTI. You will recall that the current Dual Momentum Model recommends holding 100% of the portfolio in VTI. If none of the three sector ETFs are outperforming VTI, based on PnF data, then the sector investor has these choices.
- Put all eggs in the VTI basket where one invests in small pieces of over 3,000 stocks.
- For even more diversification, invest 25% in each of VTI, VDE, VFH, and VIS.
Here is the link if you wish to check the relative strengths between VDE, VFH, VIS and VTI. When I last checked, only VFH is showing growth with respect to VTI. A conservative approach is to do the following.
- Invest 50% in VTI as recommended by the basic Dual Momentum Model.
- Invest 50% of the portfolio in VFH as it is growing faster than VTI as shown in the following PnF graph.
Now place the portfolio into “neglect” mode for another 33 calendar days.
Lee Schamp says
I was visting Antonacci’s website recently and was surprised to read on his FAQ page that he no longer recommends sector rotation.
He writes: “I no longer use sector rotation nor encourage others to use it.”
He references a blog with more data. See it as this page: http://www.optimalmomentum.com/faq.html
…Lee
HedgeHunter says
Lee,
Interesting…I’d missed this post on Antonacci’s site. I think I’ve mentioned that I’ve looked at sector rotation in the past and, while it seems logical that sector rotation may offer advantages over broader market participation, my tests have only shown a small, and not statistically significant difference in performance.
Since it seems intuitively obvious that we should be able to improve performance through sector rotation I would conclude that the simple momentum classification, on its own, is not sufficient to identify the likely best performing sectors going forward. This is one of the objectives of the new REDA system – to widen the net to include earlier identification of future high performing sectors.
David
Lowell Herr says
Lee,
Thanks for the information on sectors. I’ve never build any portfolios around sectors for this simple reason. The market covers all sectors, but sectors do not cover the entire market.
Lowell
HedgeHunter says
Lowell,
Your PnF chart uses a (User defined) $1 box size. On a (ratio) price of ~50 this is equivalent to a 2 % box size (that will look essentially the same). However, changing the box size to 1% (or $0.50) results in 3 O’s in the last column. Just a reminder that box size is an important parameter especially in conjunction with any”actions” to be taken on a column (X to O) reversal.
David