Range System

Discussion in 'Strategy Building' started by profitseer, Oct 13, 2003.

  1. It has 5 parameters, if you are interested, email me.
    Walter
     
    #11     Oct 14, 2003
  2. yea email him and he'll offer you his $4k course. you smell Walther.
     
    #12     Oct 14, 2003
  3. damir00

    damir00 Guest

    this is unbelievably easy to backtest. just pull some daily historicals from yahoo - into excel, even - and see what happens.

    if someone does actually put some work into this, i would be happy to compare/verify results.

    just PM me.
     
    #13     Oct 14, 2003
  4. DK_

    DK_

    I ran a backtest Jan 2 1997-Oct 14 2003

    Which analyzed days the SP500 tick's range > 110% average prior 30 day range

    If this tick set a new high OR low, but not both then add

    abs(
    (the max price achieved in this direction within next five days)/(breakout [high or low])
    -1)

    to a "breakout average"

    The breakout average was >0.5% 63.4043% of the time



    The same test without the 10% extra range requirement produced

    breakout average was >0.5% 61.9694% of the time



    Granted this isn't the exact method mentioned in the original post, but it's close enough to say I don't think there's a system to be had here.
     
    #14     Oct 15, 2003
  5. I exchange ideas for free. You and others with no capacity to have any useful ideas will have to pay for my time . Have you noticed that big " L " on your forehead ?

    I do not offer any course to ET members, If you find that I solicit somebody from here just post specifically who I did solicit . Otherwise you are just a pathetic liar. Big L is for loser btw but I am sure that you know that already since you are the one
     
    #15     Oct 15, 2003
  6. damir00

    damir00 Guest

    i emailed him. you're lying.
     
    #16     Oct 15, 2003
  7. ...I emailed him and sold him MY course. An easy mark. Try him.
     
    #17     Oct 15, 2003
  8. :D
     
    #18     Oct 16, 2003
  9. ram

    ram

    I emailed and PM'd Walther for his 5 parameters but got no response.:(
     
    #19     Oct 17, 2003
  10. ...due to the press of time I did not code up exactly what you suggested, but instead used an existing range breakout system as a surrogate. The rules were:

    1. NQ

    2. Don't trade before 9:30 AM ET.

    3. Long (short) when price goes above (below) a threshold

    4. Hold 'til the close.

    Whereas you talked about a threshold based on the average daily range of the market for some days lookback, this surrogate is equivalent to a 60 days lookback (the period I tested).

    The best performance was for an NQ threshold of 25. It took 8 trades and cleared $738 for each contract. However, it is trading only late in the day. Setting the threshold lower to catch earlier trades loses money due to whipsaws. I did not try to optimize it, because my experience with breakout systems for a particular time period is the following. If you have to optimize it with stops, reverses, and takes in order to make it profitable, you're looking at too long a time frame.

    I don't trade past 10:30, so I'm no expert, but my guess is that the random variations in the shape of the whole day make any kind of all day breakout system unsuitable.

    Bets regards. - Mike
     
    #20     Oct 19, 2003