Mean Absolute Deviation (MAD) Portfolio: January ETF Picks

|
 |  Includes: AGG, DBC, DDM, DOG, DXD, EEM, EFA, FXI, GSG, IAU, IEF, IEV, IOO, ITOT, IVV, IYM, IYY, JKD, JKG, JKJ, MVV, MYY, MZZ, OEF, PSQ, QID, QLD, SDS, SH, SHY, SLV, SMH, SSO, TIP, TLT, USO, VNQ, XLE, XLF, XLI, XLY
by: Jeff D. Hamann

Major economic data releases like the FHFA Housing Price Index, new home sales, durable goods, Q4 advanced GDP, and University of Michigan's consumer sentiment all come out this week. We shouldn't see major changes in the the FOMC Rate Decision (the link provides a quick look at current predictions of future rates – most predictions maintain a current fed fund rate of 0.25%), but just what does all this “news” have to do with a sensible and actionable strategic or tactical ETF portfolio strategy?

I have no idea, but for the sake of examining an asset allocation, let's assume you've suffered a stroke, or been in a coma, or have been living under a rock for the previous 252 trading days. You've just woken up, you know nothing about economics, market timing, asset allocation, or the “(mis)behavior of markets” (sorry Benoit). You call your advisor to get the current status report on your retirement portfolio, and get some report about large-cap/small-cap/blah, blah, blah... You hang up the phone dazed and confused. You're not sure if your advisor has done a good job, a poor job, or an outstanding job. You want to get some idea of what a basic portfolio that achieved your target return of 10%, gave you the smoothest ride you could get, and maintained your wealth, right? What happens next?

First, to double check your advisor, you need to gather up a universe of stuff from which you should choose your portfolio's individual assets. You want multiple asset classes and possibly a few choices within each asset class.

To make the task more simple, I'm including mostly indexed ETFs from the iShares and ProShares families of ETFs and a few from elsewhere in the bigger universe at large. The asset classes I'm including cover the major classes (Bonds, US Equities, Global Equities, Commodities, Short/Leveraged Assets, and Sector Rotation Assets). This is actually a subset of the list of ETFs provided by masterdata.com. The universe I'm using contains the following assets:

  1. Fixed-Income/Bond Assets (TIP, SHY, IEF, TLT, AGG)
  2. Commodities (IAU, SLV, GSG, BDC, USO)
  3. US Equities (OEF, IVV, ISI, IYY, JKD, JKG, JKJ)
  4. Global Equities/Emerging Markets (FXI, EFA, EEM, IEV, IOO)
  5. Inversely Correlated Assets (i.e. Shorts) (DOG, SH, PSQ, MYY)
  6. Leveraged ETFs (i.e. Longs, Double Longs, etc.) (DDM, SSO, QLD, MVV)
  7. US Equity Sectors (XLF, IYM, XLE, VNQ, XLI, XLY, SMH)

Second, you want a portfolio that will meet the challenges of today's markets and not be built upon anecdotal rules like “have your age in bonds.” Using this universe and about a year's worth of data, the portfolio results should confirm or dispute these types of rules, but that's not your goal. Based on your experience, you probably expect the market to correct 2-4 times per year and so, you want to build a basic portfolio that can handle these corrections, preserve your capital, and position you well for your next move. This means that the portfolio will most likely contain negatively correlated assets (i.e. shorts).

Today, I won't focus on pair-trades or options or other more sophisticated constructs for generating solid returns with minimum volatility. Maybe later, but for now, let's focus on the fundamentals. Today, let's examine the minumum mean absolute deviation portfolio or MAD portfolio.

Third, you'll probably want to add some constraints to limit the maximum number of included assets (e.g. 5) and limit asset class allocations for things like bonds (no more than 1 bond asset, no more than 60% bonds), sector rotation (max 2, max 20%), short/leverage (max 1, max 10%), and commodities (max 1, max 10%), with any of the remaining assets in the universe able to assume up to 100% of the portfolio, which I hope never happens.

Here are the results for today:

Exchange Traded Fund Name

Return(1)

Standard Deviation(1)

Weight

RSI (14)

iShares Barclays Aggregate (NYSEARCA:AGG)

4.95%

7.14%

59.65%

54.21

iShares Silver Trust (NYSEARCA:SLV)

63.56%

37.10%

5.86%

33.48

iShares Morningstar Mid Core Index Fund (NYSEARCA:JKG)

32.75%

29.06%

6.07%

59.71

ProShares UltraShort S&P500 (NYSEARCA:SDS)

-33.26%

52.11%

9.12%

36.28

SPDR Consumer Discretionary Select Sector Fund (NYSEARCA:XLY)

34.55%

29.71%

19.30%

54.32

Click to enlarge

The weights are the amount or proportion of capital you would allocate to each of the ETFs to obtain the minimum absolute deviation portfolio for the previous year if you had rebalanced to this portfolio 1 year ago. Some basic disclaimers:

  1. Performance results are annualized based on 252 trading days' worth of daily returns. Other analysis periods will yield different results. Dividends are not included in returns.
  2. Weights are not dependent on the initial investment.
  3. An RSI indicator of below 30 means the asset is “oversold”. Conversely, an RSI value over 70, often means that the asset is “overbought”. In the middle, is a neutral signal; watch closely.

If you've been following the previous posts, you'll note that these results are incompatible with last week's results as I've changed the assumptions for the number of trading days per year (250 to 252), the number of days' worth of data I've included (200 to 252 days), and updated, or upgraded, the constraints I've placed on the asset classes themselves.

Here's the historical performance for the previous year (click to enlarge):

Click to enlarge

Here's today's summary:

  1. The bond pick is AGG, which has a lower annualized return when compared to the previous bond picks: IEF and SHY. This might reflect a shift from government backed securities to corporate bonds, as the consequences of Quantitative Easing Part 2 (serious explanation here, fun explanation here) and the possibility of sovereign defaults (including state and local governments) might make selecting an indexed bond asset difficult over the next few months. For now, it looks like corporate investment grade bonds are the best bet.
  2. The commodities pick remains SLV, which appears to be declining from the previous commodity pick from the second week of January. The weight declined from 7.8% to 5.9%, which may be an artifact of its recent performance. It looks a little “toppy” and it might be time for a little profit taking. Might be.
  3. The US Equities pick (JKG) still looks like a solid performer, and the capital allocation change between the weight from the January 5, 2011 portfolio (8.4%) to today is 6%, which combined with the neutral Relative Strength Index (RSI) of 59 (neutral) suggests that this is still a good value for the short term.
  4. The short/leveraged position (SDS) has been remaining steady at about 9% for the past three weeks. The correction has not happened yet; and it most likely will be televised. Stay tuned.
  5. The sector rotation pick (XLY) gets about 20% of the investible capital. The sector rotation category picks seem to be changing each week. I would caution against diving into any portfolio that changes assets so frequently as this is not intended to be a market timing problem. We're looking for a tactical allocation pick. Stay tuned as the new model generates a few weeks' worth of portfolios.

The resulting portfolio, which again, is always backwards looking, has yielded the following results:

  • Target Portfolio Return/Standard Deviation (annualized) = 10.00%/5.70% - which, when compared to the benchmark I've been using (S&P500);
  • Benchmark (S&P500) Return/Standard Deviation (annualized) = 19.54%/17.69% - which has 32% of the volatility, with a nice consistent return of 10 percent annualized.

I know I said last week that I would examine the tracking error for our portfolio. I didn't do that. Sorry. I wanted to get today's standard results up so that I could focus on answering that question better next time.

Disclosure: I am long SLV.