Looking for "Turtle Soup"-system-formula for TS2000i

Discussion in 'Strategy Development' started by Jana, Nov 12, 2002.

  1. Jana

    Jana

    Thanks
     
  2. Shankar

    Shankar

    Is this the description of it?

    The TurtleSoup + 1 pattern from Connors/Raschke (Streetsmarts), buy only. The entry is defined precisely, however the exit is specified somewhat loosely. The entry tries to capture false breakouts, here: a new 20-day low that was preceded by a previous 20 day low, that was at least 3 days ago.

    Since I had a good trading day today and lots of time I did some research for you. :)

    I tested the above system on the Dow 30 stocks for the past 5 years. Starting capital was $100.000 and I used 10 % of the capital for every signal (as long as cash was available)

    PHP:

      All Trades Long Trades Short Trades Buy 
    Hold 
    Starting Capital 
    $100,000.00 $100,000.00 $100,000.00 $100,000.00 
    Ending Capital 
    $78,295.40 $78,295.40 $100,000.00 $113,037.45 
    Net Profit 
    $-21,734.60 $-21,734.60 $0.00 $13,037.45 
    Net Profit 
    % -21.70% -21.700.0013.04
    Exposure 56.9556.950.00100.01
    Risk Adj Return -38.17 -38.17 0.00 13.04 
    All Trades 731 731 0 30 
    Avg Profit
    /Loss $-29.73 $-29.73 $0.00 $434.58 
    Avg Profit
    /Loss % -0.26% -0.260.0012.93
    Avg Bars Held 10.06 10.06 0.00 1,349.00 
    Winning Trades 172 172 0 14 
    Gross Profit 
    $215,392.56 $215,392.56 $0.00 $30,151.35 
    Avg Profit 
    $1,252.28 $1,252.28 $0.00 $2,153.67 
    Avg Profit 
    12.0512.050.0064.51
    Avg Bars Held 21.57 21.57 0.00 1,349.00 
    Max Consecutive 6 6 0 3 
    Losing Trades 559 559 0 16 
    Gross Loss 
    $-237,127.13 $-237,127.13 $0.00 $-17,113.90 
    Avg Loss 
    $-424.20 $-424.20 $0.00 $-1,069.62 
    Avg Loss 
    % -4.05% -4.050.00% -32.21
    Avg Bars Held 6.52 6.52 0.00 1,349.00 
    Max Consecutive 38 38 0 4 
    Max Drawdown 
    -38.30% -38.300.00% -47.12
    Recovery Factor 0.47 0.47 0.00 0.15 
    Profit Factor 0.91 0.91 0.00 1.76 
    Payoff Ratio 2.97 2.97 0.00 2.00 
    Standard Error 
    $7,844.78 $7,844.78 $0.00 $21,066.14 
    Risk Reward Ratio 
    -0.276 -0.276 0.000 0.209 
    Sharpe Ratio 
    -3.15 -3.15 0.00 0.36 

    Hope the number make sense to you. All in all the system was not very good on stocks but it may be for futures? Sorry. I don't have the EL code.
     
  3. Jana

    Jana

    !!!
     
  4. Jana

    Jana

    ?
     
  5. a bit of history---- the turtle trading system was christened "turtle soup" after several large followers of the system blew up. :eek:

    surf:confused: :eek:
     
  6. ...in the add-on disk for STREET SMARTS (requires some updating for proper use with TS6).

    http://www.tradingmarkets.com/galleria.site/software/main/products.cfm?full=1&id=5043

    Shankar isn't specific at all about whatever exit or other criteria he used. Nor do I understand why he would try a long-only version. Anyway, I recall having had much better testing results just using the canned versions of Turtle Soup and Turtle Soup + 1. I don't trade the system as such, but I believe the concept has some validity (similar to other failed-breakout or test-of-top strategies such as Trader Vic's 2b).

    For those interested in learning EasyLanguage, by the way, the add-on provides some very useful samples.
     
  7. and I would think that the Turtle Soup strategies would work best on stocks and other vehicles that, unlike the Dow30, more typically attract fast-money breakout plays and players... Anyway, it may be worth considering.
     
  8. hi all,

    i never used turtle-soup in past but i will look for it :)

    the el-code doesnt seem to be very difficult.

    i would try to create and post this code on daily based charts if anyone is interested. (for ts6 )

    just give me post.
    ill try on saturday.
     
  9. Here is the code in WS (it might help you guys for ELA since both languages are Pascal based):

    var Bar, i: integer;
    var xStopLevel, PrevLowPrice: float;

    xStopLevel := 0.0;
    InstallTrailingStop( 10, 70 );
    InstallProfitTarget(15);
    InstallStopLoss(7);
    for Bar := 21 to BarCount() - 1 do
    begin
    ApplyAutoStops( Bar );
    if LastPositionActive() then
    begin
    if PositionLong( LastPosition() ) then
    CoverAtStop( Bar + 1, xStopLevel, LastPosition(), '' )
    end
    else
    begin
    if ( PriceLow( Bar) <= Lowest( Bar, #Low ,20 ) ) AND
    ( PriceLow(Bar -1) > Lowest( Bar -1, #Low, 20) ) AND
    ( PriceLow(Bar -2) > Lowest( Bar -2, #Low, 20) )
    then begin
    { search previous 20 - day low }
    PrevLowPrice := PriceLow(Bar);
    for i := 3 to 20 do
    begin
    if PriceLow(Bar-i) <= Lowest(Bar-i,#Low,20) then
    begin
    PrevLowPrice := PriceLow(Bar-i);
    i := 20;
    end;
    end;
    BuyAtStop(Bar+1, PrevLowPrice, '');
    xStopLevel := 0.99 * PriceLow(Bar);
    if PriceLow(Bar+1) < PriceLow(Bar) then
    xStopLevel := 0.99 * PriceLow(Bar+1);
    end;
    end;
    end;

    To all:

    Before you go and trade it, go and test it on a portfolio of stocks. The results have NOT been very encouraging as you can see.
    There are many variations on turtles on the WL site you can test them there for free and you DONT even need to code them! And do a portfolio test in the Simulator section. And here is a quote from the TS site:

    "That way, you'll always know your trading strategy's simulated historical performance before you're ready to implement it."

    Regards