Example Trading System

Discussion in 'Strategy Building' started by sidinuk, Jan 17, 2005.

  1. Well Sidinuk I must say that your article posting is great. Sure there are some contentious issues with the system but I understand your motivation was really to show a PROCESS rather than a SYSTEM - bit like teaching someone to fish rather than giving them a fish.

    So I am curious about your initial research described in the first, second and subsequent tables of the document. If you used EXCEL can you provide some information on how you derived your numbers? Specifically how did define the high/low and over what period? Was it the high for 60 minutes then 90 minutes or did you look at a days worth of 15 minute bars, note the high and low and the times they were defined? Also curious about the formula for the percentage of the day's range over 15 minute bars?

    I am not a very logical person (can see a pattern but can't describe it in code) so if I am asking a silly question please forgive me

    Thanks in advance
     
    #351     Oct 31, 2005
  2. sidinuk

    sidinuk

    alwayslearning,

    Thanks very much for your comments.

    1. I noted for each day at what time the high and low occurred and took the time of the first one. Then spit the day into 15 minute units and made a table like that. We know then that if we trade a break after this time then this is the second extreme for the day, ie we can't be stomped out if our stop is the other side of the range.

    2. We then calculate the percentage of the days range that is already in place by the breakout time to give us a risk/maximum reward ratio. The average win amount.

    3. Once we know what our percentage of trades not stopped out and our maximum average reward we can use the formula detailed here to calculate the most effective opening range breakout period.
     
    #352     Oct 31, 2005
  3. Has something changed?

    I see a couple posts back that someone refers to October 24 on the website showing a trade for YM but when I go to the website I only see data through 9/30 ?

    Thanks.
     
    #353     Oct 31, 2005
  4. Hittfeld

    Hittfeld

    Sidinuk, are you ok? Any further updates to be expected here - or on your website?

    Thanks

    Hittfeld
     
    #354     Dec 17, 2005
  5. sidinuk

    sidinuk

    Hittfeld, thanks for your concern! Yes I'm fine, unfortunately due to other commitments I haven't been able to devote so much time to updating the site recently. Hopefully I can give it some time this weekend.
     
    #355     Dec 17, 2005
  6. volaray

    volaray

    Excellent. I have actually been trading the system with a very small account via spreadbets. It has made me £500 in a couple of months.
     
    #356     Dec 17, 2005
  7. Cheese

    Cheese

    This 'Example' strategy does give some raw recognition to the gyrations of the Dow but the creators do not understand the large imprecision in their model and the potential for losses it can incur.

    Interesting, nevertheless.
    :)
     
    #357     Dec 17, 2005
  8. According to their backtested results and bearing in mind backtesting is one thing, doing is another. The Spreadbet based on 4pts cost for the DOW would knock off between 30-45% of the total profit. If its YM you have to knock off the commission and any slippage.
     
    #358     Dec 17, 2005
  9. Cheese

    Cheese

    And of course there are many times (Dow) that the day high or day low is at or a few points off the Open. If you have no means of validly recognizing or predicting that eventuality it does not help the 'Example' strategy to work other than either marginally or not at all.

    There is a good, albeit a primitively understood, idea in the whole concept. It is far far short of a finished plan. Also it appears to use available raw data and is only on a backtested basis.
     
    #359     Dec 17, 2005
  10. Cheese, I don't understand your criticism.

    As with all mechanical strategies, the trader using this one assumes that future results for the YM will be similar to the backtested ones - at least concerning the edge this system exploits. If the backtest shows that there's a significant chance YM will move further in the direction of the opening range outbreak towards close, why not try to make profits on this behaviour? I call this a valid prediction with a positive expectation value. The author has tried to implement other means, trade filters for example, but they didn't improve the results for YM. So here's something simple that works. What's wrong with that?

    By the way, I'm actively trading this system for months.
     
    #360     Dec 22, 2005