Seeking Alpha

varan's  Instablog

Send Message
Individual investor.
  • Simple Monthly Strategy Based On Polygamous Paired Switching 26 comments
    Mar 14, 2013 1:07 PM | about stocks: IJS, IJT, IJJ, IJK, IVW, IVE, IVV, EFA, EEM, TLT, LQD, IYR

    With Fidelity having just substantially increased the number of ETFs that can be traded free of commission, the following monthly trading strategy may be quite worthwhile to explore.

    Consider six baskets of various styles and markets together with fixed income and real estate ETFs:

    1. Small Cap Value and Small Cap Growth IJS, IJT, TLT, LQD, IYR
    2. Mid Cap Value and Mid Cap Growth IJJ, IJK, TLT, LQD, IYR
    3. Large Cap Value and Large Cap Growth IVW, IVE, TLT, LQD, IYR
    4. MSCI EFA and US Large Cap EFA, IVV, TLT, LQD, IYR
    5. Emerging Market and US Large Cap EEM, IVV, TLT, LQD, IYR
    6. Latin America and US Large Cap ILF, IVV, TLT, LQD, IYR

    Basically the idea is to let the data tell you which style/market/asset class is ascendant at a given time.

    At the beginning of each month, for each of these baskets, invest equal amount in the ETF whose performance was the best during the prior month.

    The results for 2003-2013 (YTD for 2013):

    CAGR 18.3%

    Sharpe .92

    Kelly Fraction .43

    Maximum Monthly Drawdown 17%

    Annual Mean Return 19%

    95% Lower bound on Annual Mean Return 9.1%

    The equity curve is shown below.

    (click to enlarge)

    Disclosure: I am long IJS, IVV, IVE, LQD, IYR.

Back To varan's Instablog HomePage »

Instablogs are blogs which are instantly set up and networked within the Seeking Alpha community. Instablog posts are not selected, edited or screened by Seeking Alpha editors, in contrast to contributors' articles.

Comments (26)
Track new comments
  • Nice work, as always, Varan. This has some similarities to Gary Antonacci's work on dual momentum.


    Monthly drawdown is a bit high for my liking, although still lower than the CAGR. So you are rewarded for the risk.
    14 Mar 2013, 02:19 PM Reply Like
  • Author’s reply » Thanks.


    My name for this strategy is Polygamous Paired Switching, as it is obviously an extension of the original paired-switching strategy wherein you pair two nominally negatively correlated assets and switch between the two in order to capture the returns during the inevitable intervals where one does better than the other. I have changed the title accordingly.
    14 Mar 2013, 02:23 PM Reply Like
  • Varan, I apologize for being a dolt, but just to make sure I understand are you suggesting you buy ONE etf from each of the 6 categories. Or both choices from inside one of the 6 categories? Thanks. jcw
    14 Mar 2013, 05:49 PM Reply Like
  • Author’s reply » No problem.


    Just one and only one from each category.
    14 Mar 2013, 06:06 PM Reply Like
  • Thank you kindly. I always find your writing to be quite interesting. Thanks for being such an excellent contributor.
    14 Mar 2013, 08:36 PM Reply Like
  • Hi Varan
    Looks like you must definitely wait 31 days between balancing to avoid charges. Doesnt invalidate anything you wrote above, just means doing it "at the beginning of each month" would be relatively expensive.



    best regards
    15 Mar 2013, 12:24 PM Reply Like
  • Author’s reply » thanks. bummer. shouldn't make much of a difference but has to be investigated.
    15 Mar 2013, 12:27 PM Reply Like
  • Same problem with TD Ameritrade free etf list. I look at the monthly holdings and see what changes. Typically, you'll find a lot of carryover from month to month, so you'd then meet the 30 day holding period. It won't eliminate transx fees, but should reduce them signficantly.
    18 Mar 2013, 11:12 AM Reply Like
  • Author’s reply » It works almost equally well with holding period of three months and evaluation period of three months.


    For 2003:2013 (In parentheses are the corresponding measures for SPY, BRK-A, and yearly updated 60/40 SPY/TLT respectively).


    CAGR 19.2% (7.3% 7.0% 8.75%)


    Sharpe .92 (.35 .3 .59)


    Kelly Fraction .36 (.24 .19 .38)


    Maximum Monthly Drawdown 19.8% (50.8% 44.5% 24.6%)


    Annual Mean Return 20.9% (9% 9.1% 9.2%)


    95% Lower bound on Annual Mean Return 9.4% (-.17% -1.5% 3.8%)


    This method appears to be quite robust.
    18 Mar 2013, 12:13 PM Reply Like
  • Good news, Varan.


    Any chance your system can evaluate an every other month evaluation (with a lookback of 1,2, or 3 months - probably doesn't make too much difference). I find it difficult to find a place where I can simulate and every other month evaluation. Everyone is focused on the monthly or quarterly - I think every 2 months might be interesting.


    Although with monthly and quarterly working quite well, I cannot see how every two months wouldn't work quite well, also.
    18 Mar 2013, 01:32 PM Reply Like
  • Author’s reply » With two months of holding period, the only evaluation period that works almost as well as the two cases that I have described is three months.
    18 Mar 2013, 02:30 PM Reply Like
  • But 3 months evaluation is better?
    18 Mar 2013, 03:08 PM Reply Like
  • varan, love your articles and posts. between you and lowell herr i'm viewing my investment planning differently. thanks so much. mjp
    16 Mar 2013, 09:27 PM Reply Like
  • Author’s reply » thank you.
    16 Mar 2013, 11:41 PM Reply Like
  • varan any thought of throwing some short products in the mix ie sh , dog...................... not,reasoning please........ thanks again for your writings
    17 Mar 2013, 11:26 AM Reply Like
  • Author’s reply » Too hard to design good periodic strategies for the inverses, as the number of down periods is much smaller than the number of up periods over any sufficiently long time periods, a circumstance that we should be thankful for, as otherwise the market will not exist, at least for individuals.


    (Of course that may not be the case for daily trading or HFT, but my experience with HFT suggests that such trading is not appropriate for portfolios less than tens of millions of dollars, and if my portfolio was that large, I would be completely out of the market, and out of SA.)
    18 Mar 2013, 12:31 PM Reply Like
  • Hi Varan,
    I like your ideas for utilizing Fidelity's expanded commission-free ETF list. Nice post!


    I'm a bit curious about including IYR in each group. I doubt that IYR would be chosen in all groups during the same month very often (perhaps never), but it seems a little risky should that happen. Did your back test show that happening any?




    4 Apr 2013, 02:19 PM Reply Like
  • Author’s reply » I have not checked that, but I agree that this situation may arise. I will take a look.


    4 Apr 2013, 02:49 PM Reply Like
  • I did a little back testing and yes you had 100% in IYR during 2008.


    Started a live run with paper money and for the period Jan 24, 2014 to Feb 24, 2014 you had 100% in IYR going forward to March 24th.


    I can't do that one!


    Maybe I am doing something wrong.
    26 Feb, 05:59 PM Reply Like
  • Author’s reply » Using 3 months of evaluation and holding periods, I do not see IYR in any of the six cases in 2008. It is however in TLT for most of the year except in the July to September quarter. That is indeed problematic, and in such a situation you will have to make your judgment.


    In 2014, it has been in IJK, IJT and IVW, and IVV in all the other three cases. The YTD return for the portfolio that is equally weighted in all the six components is 1.43%.


    Just for record, for Jan. 2 2014 (first trading day) I use the returns on the basis of adjusted closing prices on the last trading days of 9/2013 and 12/2013.
    26 Feb, 08:07 PM Reply Like
  • Thank you so much for the fast reply. Using three months to look back & using ETF replay I did get your results. But, if you look at a start date of Feb. 27th 2014 and look back three months you have 100% in IYR!


    Don't think I could do it with real money?
    27 Feb, 09:24 AM Reply Like
  • Author’s reply » I agree.


    In this case one has to use his judgment on how much to put in a single asset.


    27 Feb, 10:43 AM Reply Like
  • Author’s reply » YTD Performance:


    Quarterly updated: 12.65%
    Monthly updated: 5.8%


    We will retire the monthly strategy anyway as that entails transaction costs.


    The quarterly one has performed satisfactorily so far.
    21 Jul 2013, 10:14 PM Reply Like
  • Thanks, Varan, for the instablog. It seems a very good strategy. Do you think it might be improved including TIPS, just in case there is a bear market in bonds?
    30 Aug 2013, 05:32 PM Reply Like
  • Author’s reply » Equal amount invested in each of the six strategies with update every three months based on the prior three months of data would have returned 19.9% in 2013.


    The one containing EEM performed the worst, yielding essentially nothing.


    Overall, during 2003-2013, the updated statistics for this strategy


    CAGR 19.7%
    Sharpe 0.96
    Sortino 1.84
    Beta .44
    Alpha 1.1% (Monthly)
    Maximum Draw-down 19.8%
    Number of years of losses: 0
    2008 return: 2.6%
    5 Jan, 06:25 PM Reply Like
  • Thanks for the update.
    6 Jan, 12:45 PM Reply Like
Full index of posts »
Latest Followers


More »

Latest Comments

Instablogs are Seeking Alpha's free blogging platform customized for finance, with instant set up and exposure to millions of readers interested in the financial markets. Publish your own instablog in minutes.