Seeking Alpha

varan's  Instablog

Send Message
Individual investor.
  • 2011 was a good year for paired-switching. 24 comments
    Jan 2, 2012 8:33 PM | about stocks: VTI, EFA, EEM, IEZ, USO, GLD
    The strategy of pairing an equity with TLT and holding a position in one of the pair for 13 weeks depending on which one did better in the prior 12 weeks did not do as poorly as the market in 2011.

    Here are the returns for some equities paired with TLT for the year, together with the annualized growth rate for the pairs traded using the aforementioned strategy during the last three and five years:

    ETF  One Year (2011)  Three Years CAR  Five Years CAR
    VTI  38%  23%  23%
    EFA  37%  29%  28%
    EEM  32%  31%  31%
    IEZ  55%  45%  23%
    USO  29%  7%  15%
    GLD  45%  28%  26%

    Here are the ranking periods and investment periods for the four trades for the year:

    Investment Period  Ranking Period
    1/7/11 to 4/8/11  10/8/10 to 12/31/10
    4/8/11 to 7/8/11  1/7/11 to 4/1/11
    7/8/11 to 10/7/11  4/8/11 to 7/1/11
    10/7/11 to 12/30/11  7/8/11 to 9/30/11

    All the pairs were in TLT during the third and fourth quarter, and so the differences in the returns are only due to the differences in performance of the equities in the first six months.

    The compund annual growth rate for this strategy during the period 2005-2011 is 21%.

    The 2005-2011 results can be used to design a buy and hold portfolio of this basket on the basis of the percentage of time that the strategy was invested in each of the components. The weights of the various ETFs in this portfolio are:

    ETF  Weight for the buy and hold portfolio
    VTI  9%
    EFA  9%
    EEM  9%
    IEZ  8%
    USO  9%
    GLD  10%
    TLT  46%

    During 2005-2011, this buy and hold portfolio yielded 10% CAGR, with only one year of loss, which was 6% in 2008.Thus there are two ways to use the results: (a) use paired switching to update the portfolio every 13 weeks, or (b) invest in a buy and hold portfolio according to the allocation in the previous table, and re-balance every year.

    Obviously 2012 is another year, and there is no guarantee that the somewhat acceptable performance will be repeated.

    This is not investment advice, but just a summary of results of backtesting for the trading strategy described above.
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 (24)
Track new comments
  • Thanks Varan, very interesting results. I've been pondering paired switching for the last few months, and am going to set up a pair today or tomorrow with approx 30% or my portfolio, just need to decide which ETF to choose as opposed to TLT, since obviously that's the one I'd be buying for the next 13 weeks. Wish me luck.
    4 Jan 2012, 09:08 AM Reply Like
  • BTW, I don't see FSESX in your results (you had mentioned it as an option in a comment recently). That's the one I was thinking about using. Thoughts?
    4 Jan 2012, 09:16 AM Reply Like
  • Author’s reply » Thank you.


    FSESX is for the same sector as IEZ and yields similar returns. I used IEZ since it is an ETF and there are no limitations on holding period for it as is the case for FSESX.
    4 Jan 2012, 09:56 AM Reply Like
  • What was the 1 yr return, 3 and 5 CAR for the SPY / TLT pair mentioned previously ?
    6 Jan 2012, 06:00 PM Reply Like
  • What was the 2011 return, 3 and 5 CAR for the SPY / TLT pair mentioned in the previous couple months?
    6 Jan 2012, 06:00 PM Reply Like
  • Author’s reply » For the SPY/TLT pair, back testing yields the following returns (%):


    2003 10.9
    2004 7.6
    2005 10.4
    2006 6.7
    2007 20.9
    2008 26.2
    2009 20.7
    2010 8.9
    2011 37.9
    NET 3.9
    ANNUAL % 16.3


    1 Year 38%
    3 Years 22%
    5 Years 22%
    6 Jan 2012, 06:27 PM Reply Like
  • By my accounting, it looks like SPY returned 6.6% and QLD returned 25.3 % from 1/3/12 - the open of today ( switching into TLT )
    2 Jul 2012, 06:49 PM Reply Like
  • Compelling work ! ... Thanks ... Wish there was historical testing earlier than 2003 .


    In running some analysis, I'm going to use the SPX / TLT spread, substitue the double SP 500 ETF symbol SSO for SPY, and use a different entry rule ( other than the 1st trading day of the new quarter that applies with the normal strategy ) ... While this rule "reduced" profitability when used with SPY, it should help with a more confident entry when using SSO as we would aim for a reduction in drawdown while using a leveraged vehicle .
    In assuming that calculating the 13 week ROC of the SPY / TLT pair has signalled a SPY entry / outperformance on Jan 3rd, I will alert you to my SSO entry when the rule triggers.
    Past results are:


    alternate trading rule entries using SSO ( TLT trades not shown )
    8/11/03 -7/6/04 = +21% vs. 15% (SPY)
    1/18/05 - 4/4/05 = + 1.6% vs. -3.2 %
    10/10/05 - 7/3/06 = +9% vs. 3.4%
    3/5/07 10/1/07 = +15% vs. 7.7%
    6/29/09 - 7/6/10 = + 32% vs. 22.4%
    11/29/10 - 7/5/11 = + 20.1% vs. 17%


    7 Jan 2012, 04:50 AM Reply Like
  • Jstr, interesting idea. Can you share your spreadsheet testing SSO as your alternate strategy? What is your different entry rule?
    17 Jan 2012, 05:59 PM Reply Like
  • The entry rule for SSO has triggered for the open 3/19/12 .... Further analysis has shown that a positional split between SSO and QLD provides an added boost to returns ...
    18 Mar 2012, 08:45 PM Reply Like
  • Author’s reply » I get the following results for the SSO/TLT pair (12 week evaluation period, 13 week investment period):


    2007 27.45%
    2008 26.2%
    2009 39.9%
    2010 9.2%
    2011 44.7%
    2012 21.1%
    NET 4.3
    ANNUAL % 32.3
    Max drawdown % -19.4


    This pair has been in SSO since early this year, and has yielded 21% as shown above.


    If you do not buy TLT when that is indicated, but stay out of the market, the results are:


    2007 18.3%
    2008 0%
    2009 61.0%
    2010 2.9%
    2011 11.9%
    2012 21.1%
    NET 2.7
    ANNUAL 20.6%


    This approach would have kept you in the market only 50% of the time.


    By the way thanks for pointing out SSO as one of the ETFs to look at. If one has the stomach to trade this 2X SP500 etf (since the losses will be multiplied two fold too), it might be quite profitable. I am not quite sure I do. However, this also raises the possibility of just buying exactly in the money call options for SSO when the strategy suggests buying SSO. That obviously will further increase the risk.


    Since there is a lot of talk of TLT imploding, it is also possible to pair SPY (or SSO) with AGG. The main point is that the equities that are paired must have returns that are negatively correlated.
    18 Mar 2012, 10:42 PM Reply Like
  • Entered QLD @ 115.58 and SSO @ 58.08
    19 Mar 2012, 11:47 AM Reply Like
  • Here's the spreadsheet
    22 Mar 2012, 12:49 PM Reply Like
  • Author’s reply » Why did you wait? The SSO/TLT has been in SSO since the beginning of the year, and has returned 23%. Do you have some further filtering mechanism to refine the entry/exit points?


    26 Mar 2012, 09:59 PM Reply Like
  • I did analysis back to 2000 ( using VUSTX as proxy for 30 yr. bond vehicle ). spreadsheet:


    Over the years I've utilized "price pullback" disciplines in certain cases in order to take positions on stocks, SP futures, and ETF's ( especially leveraged ETF's). Upon analyzing the SPY/TLT pair trades, I utilized a discipline using the commodity channel index. Comparing all of the normal entry signals on the spreadsheet with the CCI pullback signals show approximately the same returns (with the exception of the last 2 ). Being that leverage is used, a later entry will most likely "catch up" to a normal entry unleveraged position's performance as long as there is more future price appreciation . The dangers that can surface are 1) with a signal such as happened in 7/9/2001 where the loss could be magnified 2) the market goes sideways and the leveraged vehicle loses "premium" ....
    27 Mar 2012, 07:59 PM Reply Like
  • Varan, can you please share the maximum draw-down for 2003/2011 for all of the pairs that you mention? Thanks.
    17 Jan 2012, 05:59 PM Reply Like
  • Author’s reply » VTI 2003-2011 -9% (Quarterly max drawdown)
    0 Number of years of losses
    16.9% CAGR


    EFA 2003-2011 -8.5% (Quarterly max drawdown)
    0 Number of years of losses
    21% CAGR


    EEM 2004-2011 -12% (Quarterly max drawdown)
    0 Number of years of losses
    24% CAGR


    GLD 2005-2011 -12% (Quarterly max drawdown)
    0 Number of years of losses
    22% CAGR


    IEZ 2007-2011 -57% (Quarterly max drawdown)
    1 Number of years of losses
    23% CAGR


    USO 2007-2011 -43% (Quarterly max drawdown)
    1 Number of years of losses
    15% CAGR


    The numbers for IEZ and USO can be improved quite a bit if you use previous three weeks for ranking and twelve weeks for investment.
    17 Jan 2012, 09:27 PM Reply Like
  • Varan - a question. How would a reverse momentum strategy work? For example, if you took the worst performing instead of the best performing of the pair?
    4 Feb 2012, 12:56 PM Reply Like
  • Varan, any opinion on how the reverse momentum strategy would work?
    2 Mar 2012, 09:27 AM Reply Like
  • Author’s reply » have not tried that.


    thanks for your interest.
    2 Mar 2012, 11:47 AM Reply Like
  • Thanks, Varan. Please let us know when you try it, Many people believe that the reverse momentum strategy should work even better,
    5 Mar 2012, 04:28 AM Reply Like
  • Hi Varan, you should submit these articles as premium articles. They are very interesting and well written, you could be making a few bucks for doing this research.
    17 Feb 2012, 12:20 PM Reply Like
  • Author’s reply » Thanks Rufus.


    Will do that in the near future.
    17 Feb 2012, 01:09 PM Reply Like
  • Have you tryed to" pairs" trade any of these, buy one and short the other. Just curious.
    20 Feb 2012, 01:33 AM 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.