Bot Based on Chart Patterns part II

Discussion in 'Automated Trading' started by endsongs, Aug 7, 2009.

  1. This bot has read and reacted correctly to the daytrading chart patterns during real money trades. Tick data backtesting helped tremendously. If it messes up, I'll report that. Curve fitting and optimization were not used. L is the average loss per trade. A summary of trades will be posted each weekend, unless it messes up.

    Total trades 48, expectancy 0.3118L, profit factor 1.713, total profit 14.96L, best trade 4.853L, worst -2.09L

    trades: as multiples of L (Calculated using actual prices obtained from trades with broker and not automated trigger price. Commission is not included as it is negligible. Plus, I wanted to see pure performance)

    July 28 total 1.666L
    0.401607759
    1.049148669
    0.188392057
    0.026946453

    July 29 total 0.0817L
    0.13520733
    0.591998996
    -0.510267687
    -0.135186605

    July 30 total 4.14 L
    3.659574327
    0.384714248
    0.096231032

    August 4 total 3.28L
    0.468876881
    4.853539221
    0.348829563
    -1.093618078
    -1.483240151
    0.182505414

    August 5 total -1.995L
    -0.805966304
    2.18905171
    -1.1079867
    -0.754212646
    0.22159734
    1.145089419
    -0.97891679
    -1.009232895
    -0.894820457

    August 6 total 1.09L
    -1.581873452
    -0.91740251
    1.905869223
    -0.610879593
    1.969631939
    -0.906529123
    -1.384024079
    2.840201838
    -0.712426332
    0.242854395
    -1.265967578
    0.513328583
    0.998297463

    August 7 total 6.709L
    4.08891177
    0.653786073
    -1.649574459
    1.52971327
    -2.09855544
    0.74694199
    -0.797149557
    -0.302169563
    4.536793657
     
  2. EMH guys can put this in their pipe and smoke it.
     
  3. What do you mean by EMH athlon?
     
  4. Efficient market hypothesis. Well, theories were made to be disproved or we would not be here.
     
  5. Aug10 to aug 19:

    Aug 10 : 23 trades, gain -18.13L

    Aug 11: gain 6.8245L

    -1.394032798
    2.921947432
    3.277215629
    -0.391691072
    0.934501465
    1.183998523
    -0.702975102
    0.99557646

    Aug 12: gain 0.89L

    1.224390937
    -1.949137374
    1.656745135
    0.761350268
    0.27513258
    0.845776514
    -1.106444929
    -0.744609797
    -0.077178354
    0.001844305

    aug 13 gain: -8.0823L

    -1.876904255
    -2.410856721
    1.141036986
    0.140679282
    -0.693440996
    -1.6084778
    -1.411908434
    -1.362433998

    aug 14: gain 10.00526L

    1.627696544
    6.471246987
    0.222071691
    0.636211135
    1.519389739
    0.350628401
    -0.231260864
    -0.926204842
    0.335480744

    aug 17 gain: 0.23425L
    0.773525329
    0.66302171
    -0.230034972
    -0.978699795
    0.271754233
    0.001844305
    -0.090495848
    -0.001844305
    -0.174815677

    aug 18 gain 0.8540L
    0.440957655
    -2.207427159
    -0.662228148
    -0.738517483
    -0.686785561
    2.851202151
    0.909420804
    0.947417153

    aug 19 gain 15.476L
    10.54041665
    0.427913084
    2.52284579
    -0.915714081
    -1.071751317
    4.151503413
    -0.351296265
    0.172445572

    Summary stats: trades 131, wins 64, losers 67, largest winner 10.54L, largest loser -2.41L, expectancy: 0.1811L, profit facor 1.354, total profit 23.721L.

    Note: trades on Aug 6 and aug 10 showed that the system logic allowed over-trading. The system code was updated for trades starting Aug 11. It's clear that the common mode rejection is now working much better. I've decided to continue trading this system with real money as, although it's not close to being perfect, it's on the right track.
     
  6. Aug 20: 14 trades, gain 2.951L
     
  7. Sick. Very well done. What broker/program are you going through for this.

    I was going to offer ideas, but from the looks of it (the size of your losses), I don't think there's much more to go here. Well done.

    Commission costs being a factor.....are they covered?
     
  8. lol exactly :)
     
  9. I'm using strategy desk Athlon.
     
  10. 8/21 to 8/27:

    (8/20 updated 13 trades , 2.99L)
    8/21 14 trades , -12.3199L
    8/24 17 trades, -5.825L
    8/25 15 trades, -1.0385L
    8/26 18 trades, -0.6790L
    8/27 13 trades, 4.7723L

    Summary: 221 trades, 100wins, 121 losses, total profit: 11.62L

    Note: new automated setups were added to the system starting Aug 21. The 23 trades for these new setups have lost about 19L (!), which is much worse than the expectancy of around 0.1515 for the rest of the trades. As tests show these new setups should be profitable long term, I'll give them 5 more trading days to show profit or I will go back to the pre Aug 21 system.
     
    #10     Aug 27, 2009