Tranche Model followers will find the following table of interest as I varied some of the positions and ranking metrics. In hopes of capturing a recovering market, I reverted back to HedgeHunter’s original ROC1 and ROC2 settings of 50% and 30% respectively. Placing more emphasis or lifting the percentage from 30% to 50% for ROC1 will capture any recovery at a quicker pace. Also note the 60 and 100 trading days as well as the Offset Portfolios setting of eight (8).
Tranche Model: The settings below are set to catch a recovering market by placing more emphasis on the most recent 60 trading days. In addition to the “Rutherford 10” ETFs I added BIV, VOE, VBR, and BWX. VOE and VBR are included to capture any value anomalies that may exist and BIV and BWX add a bond component previously missing. Based on yesterday’s data, BIV shows up as a buy in the most recent two offsets. Holding BIV and SHY at this time is not a bad position to be in as both are low volatility ETFs.
If I had chosen just two (2) Portfolio Offsets the portfolio would be divided equally among BIV and SHY. Using eight offsets results in a heavy concentration in SHY. Either way, we are out of equity ETFs during this market correction.
With all the current interest in Tranche Investing, I thought a Tranche 1.6 table showcasing this list of ETFs would be useful.
Lowell,
Could you remind me of the weighting factors of the tranches used to compute the number in the “Allocations” column. I seem to remember that the more recent offset portfolios are weighted more than the later ones, but I can’t get this from the workbook tab because the relevant cells are locked.
Thanks,
John D.
John,
Newfound Research (http://www.thinknewfound.com/wp-content/uploads/2014/11/Minimizing-Timing-Luck-with-Portfolio-Tranching.pdf) suggest that equal weighting should minimize variance. Therefore equal weighting is applied to the tranches. e.g. in the above example the allocation assigned to BIV is (50+50)/(sum of all tranche allocations, 800) = 100/800 = 12.5%.
Herb & Ernie have run tests that, I believe, may increase returns, but at the expense of increased variance in those returns – the parameter we are trying to minimize in the first place – so I’ve stuck with equal weighting.
David
That should read “Herb & Ernie have run tests using weighted returns that, ……….”
David