BioComp Dakota with Swamp Technology

Discussion in 'Trading Software' started by chifai2, May 19, 2006.

Thread Status:
Not open for further replies.
  1. jalsck

    jalsck

    An equity engine based on the 'percent direction correct' performance metric is being used. The labels for the bars long and bars short stats are shown, but are not calculated. It is possible to build your own equity engine using VB6, VB.net or C# and calculate / display whatever statistics you want to. BioComp Systems have sample projects available for download.

    Kind Regards,

    James
     
    #81     Jun 2, 2008
  2. rickty

    rickty

    James,

    How do you combine these 39 signals? Is it simply a majority vote or something more involved?

    Also, I can't help thinking 39 signals is overkill. Did you actually verify the value added of all 39 signals? If so, how could you do that considering the very large number of combinations possible.

    Richard
     
    #82     Jun 2, 2008
  3. jalsck

    jalsck

    Richard,

    I do have an EST (ensemble signal trader script bot) that is based on a majority vote with adaptive thresholds. This works reasonably well. Sp_z0020 uses different combinations of signals moving forward, based on the equity acceleration (change of the change in the equity) produced by each signal. The parameters are adaptive for this one as well.

    More systems provides more diversity (assuming they are not too similar). However, adding more signals can erode hypothetical returns. I see it as a two way street. Tonight, I am doing a quick investigation into a more sophisticated way of combining the signals.

    I know there is room for improvement because using BioComp Profit with data derived from my current Dakota systems I can produce a system that performs at about 20% of perfect over a 3 year out of sample period. I would like to produce a similar equity curve on a purely walk-forward basis using Dakota.

    Kind Regards,

    James
     
    #83     Jun 2, 2008
  4. lucciri

    lucciri

    Dear James,

    yes , thank you . now I use WaveoscHarmonicdblsmaposentryfiltereqtyctrls (gasp!) for ES , with real money. I am interested in ensemble a signal, unfortunately I am not a programmer, can you suggest an easy way to start something simple . i.e just two-three signals? thank you , (sorry for my english)
     
    #84     Jun 2, 2008
  5. jalsck

    jalsck

    Signal for Tue Jun 3, 2008 MOC:
    Long

    Trading Instructions for Tue Jun 3:
    None

    Open Position:
    Long 1 ESM8 @ 1385.25

    Trading History:
    4/1/2008 Bought 1 ESM8 @ 1371.00
    4/22/2008 Sold 2 ESM8 @ 1380.25
    5/1/2008 Bought 2 ESM8 @ 1412.1
    5/7/2008 Sold 2 ESM8 @ 1394.7
    5/8/2008 Bought 2 ESM8 @ 1392.5
    5/19/2008 Sold 2 ESM8 @ 1429.3
    5/27/2008 Bought 2 ESM8 @ 1385.25

    Profit / Loss From Closed Positions:
    4/22/2008 $462.50
    5/1/2008 ($1,592.50)
    5/7/2008 ($870)
    5/8/2008 $110
    5/19/2008 $1,840
    5/27/2008 $2,202.50

    Note: This is a hypothetical trading record.

    Kind Regards,

    James
     
    #85     Jun 3, 2008
  6. jalsck

    jalsck

    Signal for Wed Jun 4, 2008 MOC:
    Short

    Trading Instructions for Wed Jun 4:
    Sell 2 ESM8 MOC

    Open Position:
    Long 1 ESM8 @ 1385.25

    Trading History:
    4/1/2008 Bought 1 ESM8 @ 1371.00
    4/22/2008 Sold 2 ESM8 @ 1380.25
    5/1/2008 Bought 2 ESM8 @ 1412.1
    5/7/2008 Sold 2 ESM8 @ 1394.7
    5/8/2008 Bought 2 ESM8 @ 1392.5
    5/19/2008 Sold 2 ESM8 @ 1429.3
    5/27/2008 Bought 2 ESM8 @ 1385.25

    Profit / Loss From Closed Positions:
    4/22/2008 $462.50
    5/1/2008 ($1,592.50)
    5/7/2008 ($870)
    5/8/2008 $110
    5/19/2008 $1,840
    5/27/2008 $2,202.50

    Note: This is a hypothetical trading record.

    Kind Regards,

    James
     
    #86     Jun 4, 2008
  7. jalsck

    jalsck

    Signal for Thu Jun 5, 2008 MOC:
    Short

    Trading Instructions for Thu Jun 5:
    None

    Open Position:
    Short 1 ESM8 @ 1377.00

    Trading History:
    4/1/2008 Bought 1 ESM8 @ 1371.00
    4/22/2008 Sold 2 ESM8 @ 1380.25
    5/1/2008 Bought 2 ESM8 @ 1412.1
    5/7/2008 Sold 2 ESM8 @ 1394.7
    5/8/2008 Bought 2 ESM8 @ 1392.5
    5/19/2008 Sold 2 ESM8 @ 1429.3
    5/27/2008 Bought 2 ESM8 @ 1385.25
    6/4/2008 Sold 2 ESM8 @ 1377.00

    Profit / Loss From Closed Positions:
    4/22/2008 $462.50
    5/1/2008 ($1,592.50)
    5/7/2008 ($870)
    5/8/2008 $110
    5/19/2008 $1,840
    5/27/2008 $2,202.50
    6/4/2008 ($412.50)

    Note: This is a hypothetical trading record.

    Kind Regards,

    James
     
    #87     Jun 5, 2008
  8. James,

    Now that the trades are starting to accumulate, would it be too much trouble to post a running total of the P/L? I'm sure many of us are adding them in our heads to keep track of how things are going overall.

    But if it's going to add significant work then forget it, I can live with what you're posting now.

    Thanks.
     
    #88     Jun 5, 2008
  9. One other comment: if you were to average 17% of perfect (per the screen shots you posted earlier), and if perfect for the ES was $170K/year, (my data from 5/24/07 to 5/23/08), then that should yield about $28.9K/year, or $2.4K/month.

    Does that sound right, or did I make a mistake somewhere?

    I think right now the swarm is yielding about $850/month, or about a third of what I'd expect. Of course, I very much realize that's on a small sample of trades. At this point, all it takes is one large trade and you're caught up. (or really in the hole! :) )
     
    #89     Jun 5, 2008
  10. I meant to say "and if 100% perfect for the ES was $170K/year" in my previous post.
     
    #90     Jun 5, 2008
Thread Status:
Not open for further replies.