E-mini SP Counter Trend Challenge

Discussion in 'Strategy Building' started by MarkBrown, Jan 8, 2008.

  1. MarkBrown

    MarkBrown

    ok i don't know who's system your asking for but on mine i figured 40.00 commissions and 100.00 slippage and that's for the e-mini sp. ya think that's enough? ;)

    mb
     
    #21     Jan 12, 2008
  2. Yes. Please post performance reports showing these figures included.

    Thanks!
    M
     
    #22     Jan 12, 2008
  3. MarkBrown

    MarkBrown

    look at the last few pictures and you will see i added 100 slip and 40 commish. i also could have chosen a dozen of so per trade profits from 400 to 1000 and the system also doesn't really care within a day when you place the trade. i think that demonstrates the robustness of counter trend trading.

    mb

    http://www.yafeu.com/pic/
     
    #23     Jan 12, 2008
  4. MB,

    Total number of trades is a little low, got anything in the thousands ?

    Thanks

    Anek
     
    #24     Jan 12, 2008
  5. MarkBrown

    MarkBrown

    yea i do i have some stuff on natural hour data with 17,000 to 20,000 bars which makes a few thousand trades. i usually make my own synthetic data series for eod and test all this stuff on some pretty crazy data prior to putting it on real data. whats neat is if you did it right the settings will be pretty much dead on. but i can tell you i do demand of myself to have as many trades as possible.

    my one minute day trading stuff makes 40 trades a day on data series testing back to i think 87 which would be 2,100,000 bars of data which the system is over 60% profitable. that's going thru a few markets transitions isn't it? btw that's also 200,000 trades which i have made about at least half in real trading.

    mb
     
    #25     Jan 12, 2008
  6. You mentioned the abracadabra words, synthetic bars, that was the key for my automation.

    Anek
     
    #26     Jan 12, 2008
  7. MarkBrown

    MarkBrown

    yea you know when you can do it on that synthetic data - you have it. mb

    i use an old program called metrodome to make my data - you can set all kinds of obstacles up in the data like spikes and all sorts of tricks. they don't sell it anymore.

    in fact i test on so much data that i only test on close data now (easier to handle all the ticks) cause if you think about it if you have a bad tick in real time data it will always be the high or the low of the bar not the close, usually anyway. just another lesson i learned the hard way. you give up some profits but its safer.
     
    #27     Jan 12, 2008
  8. rickty

    rickty

    Why's that? I can't see what synthetic bars have to do with automation. Can you please explain?

    Richard
     
    #28     Jan 12, 2008
  9. MarkBrown

    MarkBrown

    well think of it this way - we can train a monkey to perform a task to get a banana right. but take that monkey out of his lab and try and get him to do a new trick he has never done before to get a banana and guess what he will fail. so he like the system wasn't too smart in the first place it was just an illusion.

    mb
     
    #29     Jan 12, 2008
  10. Be sure to let us know once rubber meets road.

    Happy Trading!
    jj
     
    #30     Jan 12, 2008