FREE ES Trading Strategy that works

Discussion in 'Strategy Building' started by DonKee, Apr 29, 2009.

  1. ESSTrader

    ESSTrader

    Nice trades. Thanks for sharing.

    I only had one winner trade early in the day and indeed didn't get another cross of the zero so missed the massive move.

    Looks like reading the type of day we have on hand can be key to how you play it: taking first pullbacks only vs. fading range extremes vs. taking almost every pullback in a trend day.

    So far I have on my screen:

    Daily bars to check prior day's range
    TICK readings
    Total volume

    More stuff to work on :)

    I wonder if I should switch to 500 ticks as well. I've been a bit reluctant to do it because in my experience (in stocks) switching between timeframes on a regular basis can be tricky and frustrating. But it might be different in this case.
     
    #181     Jul 15, 2009
  2. DonKee

    DonKee

    I'm not switching to the 500 tick chart. I just have it up with the 1000 tick chart. As I mentioned before, the summer months are low volume and you want to use whatever chart is describing the price action in a way that gives you good entries. Today is a trending day, I want to make sure I can either enter a trade or add on to my positions on pull-backs.

    Once you get the "hang" of reading the 3/10/16, you'll be able to put up a chart of just about any market and find a time frame that is describing the pull-backs and momentum exhaustion in a way that will give you a nice edge on your entries.

    BTW: Now that we have re-tested the 26 area, I'm not going to be aggressive on getting long. At this point, I'm moving up stops to let the market take me out.
     
    #182     Jul 15, 2009
  3. Schaefer

    Schaefer

    Thanks DonKee for your efforts. For those who are interested, this method also works very well on NQ with volume bars.

    But days such as yesterday would be losing days for this kind of trend following system. You could of course, protect yourself with some filters to minimize the damage.

    Have fun,

    Schaefer
     
    #183     Jul 15, 2009
  4. lynx

    lynx

    DonKee: Looking back to my post from yesterday, I had misread yours from the 14th (confused about what day it was since I posted after midnight, yes, I'm stupid).

    I got different signals than you did. I was using the Trade Station built in MACD, with 3-10-16. Is that MACD the correct one? I know you said you wrote your own.

    Thanks.
     
    #184     Jul 15, 2009
  5. DonKee

    DonKee

    You're going to need to make sure the MACD is using SMA, not EMA.
     
    #185     Jul 15, 2009
  6. lynx

    lynx

    Yep, that's it, the standard TS MACD is using exponential averages.

    I wrote my own indicator to compute MACD with simple averages. Would you mind taking a quick look at my chart to see if it looks like what you are seeing? I don't get the exact same trades you saw on the 14th.

    Thanks for taking the time to share this. I feel that it's going to be worthwhile but I've got a ways to go before I fully get what is happening here.

    Here is my TS code if you are interested.

    Code:
    inputs: FastLength( 3 ), SlowLength( 10 ), SignalLength( 16 ) ;
    variables: vFast(0), vSlow(0), vMACD(0), vSignal(0), vHist(0);
    
    vFast = Average(Close, FastLength);
    vSlow = Average(Close, SlowLength);
    vMACD = vFast - vSlow;
    vSignal = Average(vMACD, SignalLength);
    vHist = vMACD - vSignal;
    
    Plot1( vMACD, "MACD" ) ;
    Plot2( vSignal, "Signal" ) ;
    Plot3( vHist, "Histogram");
    
    Plot4( 0, "ZeroLine" ) ;
    
     
    #186     Jul 15, 2009
  7. DonKee

    DonKee

    Try:

    inputs: Price(Close), FastLength( 3 ), SlowLength( 10 ), SignalLength( 16 ),Displaced(0);
    variables: vFast(0), vSlow(0), vDiff(0),vDiff1(0), vSignal(0);

    vFast = AverageFC (Price, FastLength);
    vSlow = AverageFC (Price, SlowLength);
    vDiff = vFast - vSlow;
    vDiff1=vFast - vSlow;
    vSignal = AverageFC (vDiff, SignalLength);


    if Displace >= 0 or CurrentBar > AbsValue( Displace ) then
    begin

    Plot1 [Displace] ( vDiff, "Diff" ) ; //Set this as a Histogram
    Plot2 [Displace] ( vDiff1, "DiffLine" ); //Set this as a Line
    Plot3 [Displace] ( vSignal, "Signal"); //Set this as a Line
    Plot4 [Displace] ( 0, "ZeroLine" ) ;

    end;


    Make sure you set "Diff" as a Histogram and Set "DiffLine" as a Line

    You can mess around with the colors, thickness, etc.

    You should be good.
     
    #187     Jul 15, 2009
  8. lynx

    lynx

    Thanks for straightening me out. It would have taken me a while to figure out that I was plotting the wrong values. :)
     
    #188     Jul 15, 2009
  9. DonKee

    DonKee

    Displaced(0) should be Displace(0)
     
    #189     Jul 16, 2009
  10. WooDoo

    WooDoo

    the truth is that this simple strategy does not work. Have tested it with about last 10 days data on ES, NQ, YM futures. comissions and slippage will eat all you profit.
    If this system based on some kind of intuition or ezoteric knowledge, you can use any other indicator for confirmation of your decisions.
    Some notes about this "simple strategy" if somebody will risk to play around it.. I would not use it in the long run.

    1. Suprisingly, Macd with SMA works better than EMA's or even other now days sophisticated MA,s
    2. Using SMA in MACD formula and 500 or 1000 tick time frames gives you a chance to stay at least on zero loss/profit if you will be lucky enough (im not sure about this in the long run)
    3. This system absolutely useles for shorting, at least at current market conditions
    4. stops 2 and 2 gives you some theoretical profit, but PT at 10 or about 1 percent is better. There are some more loss/profit target combinations, for example tight 0.005 percent stop and 0.3 profit target,

    Below is 10 days simulated ES, NQ, YM futures basket trading report, with 2/2 stops and raw code (use at your own risk:).

    Sorry for pure english
     
    #190     Jul 18, 2009