Seeking Alpha
Long only, ETF investing, portfolio strategy, momentum
Profile| Send Message| ()  

Reducing portfolio risk through the "Dynamic" plus SHY model was explained in Part 1. Is there a model that does not require an optimizer yet accomplishes the same goal of increasing the Return/Risk ratio for a portfolio? The following "Weighted Momentum Plus SHY" model does just that.

As in the previous article, one first selects a group of ETFs that provide the opportunity to construct a global portfolio. In the first part of the following example 22 ETFs are selected. The total securities tally indicates 23, but I dropped SDS, a short ETF, off the list as it may be confusing to many investors.

The performance ranking shown below is based on 9/19/2013 data. Fifty percent (50%) of the momentum ranking is tied to performance over the past three months. Another thirty percent (30%) is allocated to the six month performance and the final twenty percent (20%) comes from a volatility calculation where low volatility receives a higher score.

Instead of letting the optimizer determine the number of shares to hold in each ETF, in this example we look for ETFs that are ranked higher than SHY and among those we assign a percentage of the entire portfolio with larger percentages allocated to the top performing ETFs. In this case we allocate 15% to the top three performers, VBK, VOT, and VEA. Then decreasing percentages are assigned to ETFs as they are lower on the performance or momentum scale. These are arbitrary percentages in this example and they are not the weighted percentages assigned in the performance graph shown in the second screen shot. Think of this as a snapshot should the portfolio be up for review today.

If this portfolio were up for review, we would buy 30 (rounded) shares of VBK to fulfill the 15% weighting to this particular asset class. In the case of VTV we would sell 30 shares. For this example I am using a portfolio that is approximately $100,000 in size.


(Click to enlarge)

The Feynman Study began in late June of 2007 and ended in late June of 2013. This period begins near a market high, includes both the deep recession of 2008, and the five-year recovery since March of 2009. Instead of using an optimizer, as we did in Part 1, in the weighted momentum model different dollar weights give more preference to those ETFs with the higher performance rank when the portfolio comes up for review. While I review portfolios every 33 days, The Feynman Study updated the portfolio every quarter.

Each quarter the portfolio is made up of ETFs that perform above SHY, the cutoff ETF. In The Feynman Study there were 18 possible ETFs so there could be up to 17 ETFs included if SHY ended up as the lowest ranked ETF. In the above example we have 22 ETFs and the current ranking shows SHY at number 15. If SHY were ranked number one, the entire portfolio (100%) would be allocated to SHY.

In the six-year period of the study (re-balanced quarterly) the maximum number of ETFs held was 16 and the minimum (excluding SHY) was 1. Thus the portfolio contains a wide range of allocations depending on market conditions. This should not come as a surprise considering the bear market of 2008 and early 2009 when equities were in free-fall. This is exactly the reason for including all the bond and treasury options. One can flee to those asset classes in times of trouble, and if bonds do not provide shelter, we move assets to SHY. Another option is to leave the cash in a money market.

The performance of the "Momentum Weighted" Portfolio is shown in the following figure. Pay attention to the top and bottom graphs. The bottom graph shows the performance of Vanguard's Total Market Index Fund, VTSMX. This fund was used as a reference since it performed above VFINX, a mirror fund to the S&P 500. The top graph is the performance of the Weighted Momentum Portfolio. The Weighted Momentum Portfolio does not hold ETFs if they are performing below SHY, the cutoff ETF. Of those ETFs performing above SHY, a higher percentage is assigned to the higher performing ETFs.

In addition to improving return, risk is reduced as one can see by comparing the two graphs.


(Click to enlarge)

Key features of the Feynman "Momentum" Portfolio with SHY momentum cutoff filter are:

  • 6-year return: +110.99% or Compound Annual Growth Rate (OTCPK:CAGR) = 13.25%
  • Maximum Draw-down: -14.41%
  • Maximum Trough to Peak: 110.99%
Source: Momentum Model Increases Return While Reducing Portfolio Risk - Part 2