Marc Cohn's  Instablog

Marc Cohn
Send Message
I am a patent litigator with a background in physics and electronics. I enjoy studying quantitative, rules-based investment methods through rigorous backtesting and numerical analysis. I believe that patterns exist in the market that benefit trading -- the challenge is finding them!
  • Modified GMR Signal For August 2014: EEM 24 comments
    Aug 18, 2014 3:43 PM

    Sorry for the absence, everyone! My day job has been keeping me very busy and it will do so until late-September at least.

    For those of you who guessed EEM as August's signal, you were correct. EEM it is.

    I will try not to be remiss for next month.

    Cheers,

    M

Back To Marc Cohn'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
  • Hank Pong
    , contributor
    Comments (14) | Send Message
     
    Great to hear from you
    18 Aug 2014, 03:53 PM Reply Like
  • Algyros
    , contributor
    Comments (164) | Send Message
     
    Yes, very nice to hear from you. And thanks.
    18 Aug 2014, 04:20 PM Reply Like
  • olmendreef
    , contributor
    Comments (49) | Send Message
     
    Good to have you back Marc.
    18 Aug 2014, 05:59 PM Reply Like
  • berry
    , contributor
    Comments (169) | Send Message
     
    If Marc get's busy again ;) and/or you're interested in a "performance only" approximation of Marc's GMR, (it's been correct every month so far), you can visit etfrank.com. GlobalSSO - use the "Performance" (column 1) ranking.

     

    If you rank it utilizing "GMR" 20 day volatility (column 3 "Rank"), the results may vary slightly from Marc's. Possibly due to SSO's slightly higher volatility component or the fact that it is utilizing 20 day, rather than 6 month, volatility.
    19 Aug 2014, 04:51 AM Reply Like
  • Algyros
    , contributor
    Comments (164) | Send Message
     
    Thanks, Berry.

     

    And thank you for your site and your tests. If you don't mind, could you tell us if you've backtested Marc's version of GMR (without volatility and with the cash stop) and, if so, what your results are?
    19 Aug 2014, 09:12 AM Reply Like
  • berry
    , contributor
    Comments (169) | Send Message
     
    Hi Algyros,

     

    Without attempting to hijack Marc's posts - here's our "comparable" back-test - with the following caveats.

     

    1: It uses 67/33 performance/20 day volatility - this tests better than performance-only for most time periods. It's what we use for almost all forecasts.
    2: Includes NO cash stop. Our experience with cash stops is very mixed so we have chosen not to include one.
    3: Max Draw-down is almost 23% (Jan/Feb 2009). A cash-stop may have mitigated this, but see 2 above.
    4: Note that not all the ETFs include data for the entire period tested e.g. EDV
    5: As with all of these back-tests, we fear there is some amount of selection-bias. We are attempting to quantify whether it exists and it's possible significance and will be publishing our results and conclusion. In the meantime, please read the cautions/disclaimers on the site and don't rely on any of these systems continuing to return 40% annually into the future. This is not an attempt to discredit any of the fine work that has been presented here or by Mr Grossman. As you can see, our own test indicates a potential for 40%+. We're just not sure exactly how representative that test will be in the future or even if it was truly representative of the past.

     

    http://bit.ly/1rUg7va

     

    berry@etfrank.com
    19 Aug 2014, 04:18 PM Reply Like
  • EdwardjK
    , contributor
    Comments (197) | Send Message
     
    MCohn's modified GMR strategy has been pin place since March, 2014. Can someone post their actual YTD performance using this strategy?

     

    Thanks.
    19 Aug 2014, 09:23 AM Reply Like
  • Chuck31
    , contributor
    Comments (27) | Send Message
     
    Marc, I've enjoyed catching up on your GMR strategy the past couple of weeks. I've finally gotten an Excel spreadsheet done and it seems to be accurate. If Mark or anyone else has the chance, would you verify that the following numbers are correct?

     

    If August ended yesterday (Friday Aug 22), I'd have ILF as the September pick. Using 8/22 close prices:
    ILF 9.34%
    FEZ -7.54%
    EEM 5.82%
    SSO 7.16%
    EPP 3.90%
    EDV 5.38%

     

    sso/edv correlation at .416
    23 Aug 2014, 03:13 AM Reply Like
  • berry
    , contributor
    Comments (169) | Send Message
     
    Using performance only, I have
    SSO 10.9
    EDV 8.3
    ILF 6.2
    EPP 4.8
    EEM 4.36
    FEZ -5.3

     

    Since it's mid-month, I'm looking back 63 days.
    23 Aug 2014, 06:24 AM Reply Like
  • scarecrow93
    , contributor
    Comments (5) | Send Message
     
    I currently have:
    ILF 9.49%
    SSO 7.48%
    EEM 5.88%
    EDV 5.48%
    EPP 3.95%
    FEZ -7.36%

     

    Cash stop correlation: +.41
    23 Aug 2014, 12:54 PM Reply Like
  • IndyDoc1
    , contributor
    Comments (312) | Send Message
     
    It looks like ILF is the winner for the month of September
    29 Aug 2014, 04:44 PM Reply Like
  • Steadyprofit
    , contributor
    Comments (42) | Send Message
     
    My big question is how many people rolled right out of EEM and into ILF today? With ILF's recent rebound, I'll be the first to admit that I went to cash today and didn't buy ILF. I know this type of behavior defeats the purpose of a purely mechanical system, but I would prefer at least an intra-day pullback before buying. Won't be the last poor decision I've made if ILF keeps going up and I missed the boat. Guess that's what makes investing so interesting. Everyone enjoy your extended weekend.
    29 Aug 2014, 06:02 PM Reply Like
  • tmdoherty
    , contributor
    Comments (1171) | Send Message
     
    IndyDoc1, my spreadsheet also has ILF. ILF is certainly overextended now and ripe for a pullback, but there isn't even a hint of technical weakness, so I will still roll into that position on Tuesday.

     

    Here are the 3 month results I have:

     

    ILF: 13.96%
    SSO: 8.68%
    EDV: 7.95%
    EEM: 6.51%
    EPP: 3.69%
    FEZ: -5.94%

     

    Cash stop correlation = +0.598
    30 Aug 2014, 12:06 AM Reply Like
  • Chuck31
    , contributor
    Comments (27) | Send Message
     
    tmdoherty, my numbers exactly match yours, except for the cash stop correlation, which I have a little higher at +0.612. Since everything else is exactly the same, the correlation must be a difference in rounding.

     

    The correlation has been creeping up pretty steadily. I was wondering earlier in the week if it was going to reach the stop at 0.75.
    30 Aug 2014, 11:04 AM Reply Like
  • IndyDoc1
    , contributor
    Comments (312) | Send Message
     
    How do you calculate the Cash Stop Correlation?
    30 Aug 2014, 01:12 PM Reply Like
  • Chuck31
    , contributor
    Comments (27) | Send Message
     
    I've been calculating it the way it was described in the comments back a couple of months ago: take the monthly percentage return for SSO for each of the last four months, in this case for each of May, June, July and August. Same thing for EDV. I plug those two arrays of four numbers each into Excel's CORREL function to get the correlation.
    31 Aug 2014, 04:18 PM Reply Like
  • IndyDoc1
    , contributor
    Comments (312) | Send Message
     
    Thanks !
    1 Sep 2014, 07:06 AM Reply Like
  • nike X
    , contributor
    Comments (25) | Send Message
     
    tmdoherty, my results, including the cash correlation, match yours exactly.

     

    Korndog, good move going to cash before the close Friday with ILF clearly out in front and a three day weekend coming up. I think the chances of a geopolitical gap down at Tuesday's open are greater than the opposite happening. I would have done the same, but Friday afternoon got away from me.

     

    In general, I am constantly (sometimes painfully) reminded to stay with the mechanical (computer optimized) game plan.

     

    I will be switching EEM => ILF Tuesday morning.

     

    Regards.
    31 Aug 2014, 07:17 PM Reply Like
  • Chuck31
    , contributor
    Comments (27) | Send Message
     
    I just recalculated the correlation using the natural log of the monthly return ratio for SSO and EDV for the past four months, and now the correlation also comes to +0.598. Apparently I had been using the wrong formula that yielded almost but not quite the right results.
    31 Aug 2014, 10:55 PM Reply Like
  • tmdoherty
    , contributor
    Comments (1171) | Send Message
     
    Right Chuck31, that's the correct formula: calculate the correlation coefficient of the natural logs of 1 month returns over the most recent 4 months for EDV vs. SSO. Glad everybody agrees now.

     

    Terence
    2 Sep 2014, 01:13 AM Reply Like
  • IndyDoc1
    , contributor
    Comments (312) | Send Message
     
    EEM and ILF have 0.8 correlation. So , even if you did not switch to ILF, I think it will be OK especially ILF have higher volatility
    1 Sep 2014, 07:10 AM Reply Like
  • Hank Pong
    , contributor
    Comments (14) | Send Message
     
    Great, switch to ILF
    1 Sep 2014, 09:55 PM Reply Like
  • berry
    , contributor
    Comments (169) | Send Message
     
    For anyone who wishes to see the effects of various performance/volatility combinations in calculating a performance/volatility ranking, there's a somewhat crude demo program available for download on the etfrank site.

     

    As stated, it's still pretty crude and it's purpose is NOT for developing new record breaking strategies but rather to demonstrate how relatively small changes in parameters and timing can have a sometimes significant impact on results.

     

    Feedback is welcomed.
    15 Sep 2014, 01:29 PM Reply Like
  • berry
    , contributor
    Comments (169) | Send Message
     
    September was not a kind month for this ETF strategy.

     

    If you're contemplating the effectiveness of a momentum strategy during turbulent times and might like to look at developing some alternatives on your own, we've upgraded our utility to allow you to evaluate strategies utilizing a combination of performance and volatility via portfolios built of your selections from over 400 of the largest ETF families.

     

    As stated on the site, it IS in beta status. It's entirely self-contained and utilizes it's own database of ETF data.

     

    http://etfrank.com
    3 Oct 2014, 05:24 PM Reply Like
Full index of posts »
Latest Followers

StockTalks

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.