My automated system - thoughts

Discussion in 'Automated Trading' started by travis, Oct 24, 2007.

  1. travis

    travis

    I want to think out loud about my strategy and if anyone has any ideas about it please let me know. Also I want to clarify it to myself.

    This strategy, more or less, intermittently, makes money, but that is not the point. Let's think about why it should make money, and if it's likely to make money in the future.

    It trades futures. Forex, Indexes, Bund.

    It monitors these markets all day long, recording highs and lows, and then starts trading at 9.30 EST.

    Whatever high and low were recorded until 9.30 are considered support and resistance levels, and my system tries to catch bounces off support and resistance and near those levels (within a distance equivalent to 25% of the daily range).

    That's not it though. We have 2 moving averages (6 minute and 15 minute price averages) that have to cross each other, to trigger a trade and signal at the same time that price has reverted its trend and there is a bounce (it doesn't guarantee anything of course). Also, a certain amount of time (40 minutes) has to pass after a new low/high is recorded before a trade can start off that support/resistance level.

    These two features (moving averages and time to wait after a new high/low) are meant to avoid catching a high that is not a high, but that soon will be followed by a higher high. How do we know that price has stopped rising for good?

    Well, to begin with, if we just reached the day's high, we should wait a little (40 minutes) to see if it keeps on rising. As soon as 40 minutes have passed from the latest high, we can say that the rising has stopped for the moment. The time of the day will also help a little bit, because usually the markets are running wild in the first hour of trading, and then bounce off. So if we allow it 40 minutes after a high, that will probably be it (a higher high would follow before the end of the 40 minutes).

    The other good feature of the system - the moving averages - prevent a trade to start when price is going in the wrong direction. Say we reached a high over 40 minutes ago, but price is heading toward that same high, and may cause a higher high. As long as price is going in that direction, the 2 moving averages will not cross each other, and that will prevent trading even if the 40 minutes have passed.

    Look at the picture below.
    The green circle shows how the trade could be triggered - it's after 9.30 EST, 40 minutes have passed after the new HIGH, and there is a bearish crossover.
    The red circle shows that the trade cannot be taken and gladly because it would have failed - it's after 9.30, 40 minutes have passed after the new HIGH, and there is NOT a bearish crossover, but price keeps going higher, and we will have to wait 40 more minutes after that new HIGH.
     
    • snap.jpg
      File size:
      47.9 KB
      Views:
      396
  2. novais

    novais

    I have comments and questions below:

    >It monitors these markets all day long, recording highs and lows, and then starts trading at 9.30 EST.

    Please clarify monitoring "all day long" and then start trading at 9:30 EST. What is start time for monitoring?

    > Whatever high and low were recorded until 9.30 are considered support and resistance levels, and my system tries to catch bounces off support and resistance and near those levels (within a distance equivalent to 25% of the daily range).

    How many days you average for daily range?

    > That's not it though. We have 2 moving averages (6 minute and 15 minute price averages) that have to cross each other,

    I thought both to cross support or resistant level, rather crossing each other. I guess noise/filter will be taken care by price averaging during 6 or 15 minutes.

    > Also, a certain amount of time (40 minutes) has to pass after a new low/high is recorded before a trade can start off that support/resistance level.

    This is good approach to re-bench mark for finding next trading opportunity. If I see other side I think this is another way to restrain ATS to let it not over trade in a day :).

    How many trades (on average) on one index(or whatever) your ATS is making?

    > These two features (moving averages and time to wait after a new high/low) are meant to avoid catching a high that is not a high, but that soon will be followed by a higher high. How do we know that price has stopped rising for good?

    Do you wait 40 minutes after new high/low?

    In summary to make is simpler [for me] monitor trades in last 40 minutes to form R/S level, then when 6/15 minutes MA crosses them make trades. After making trade, wait 40 minutes to find new R/S levels

    How do you use trailing stop after entering in to a position?

    What logic do you use to add to a position?

    Thanks for sharing your ATS thoughts!
     
  3. travis

    travis

    Are you kidding me? How much are you paying me to answer all these questions? Besides, I am not going to tell you exactly how my system works. Lastly, I already told you to meet me on msn messenger so we can discuss things more easily instead of sending questions and answers back and forth.
     
  4. novais

    novais

    >Are you kidding me? How much are you paying me to answer all these questions? Besides, I am not going to tell you exactly how my system works.

    My mistake. I thought you started this thread to share thoughts about your ATS. I wasn't sure where to draw line. I take my questions back.

    > Lastly, I already told you to meet me on msn messenger so we can discuss things more easily instead of sending questions and answers back and forth.

    I attempted last week. I guess due to time difference we were not online at the same time.
     
  5. travis

    travis

    I am online during the week, between 6 pm and 10 pm European Central time. Tell me a day and an hour and we can meet, better in the weekends. It's too complicated to answer all these questions on a forum.
     
  6. travis where in europe are you,

    do u work full time?

    and can u share ur age??

    tnx
     
  7. travis

    travis

    Age, over 30. Country, Italy. Work, not full time. I work at a bank, that is why I needed an automated system.
     
  8. If you post many more pictures of your method on much more data I would be more than happy to shoot holes in your model all day long. Plus if you think you have something worth while - your mistaken.

    Successful modelers throw out much more valuable ideas than this example because they understand to give is to get. You may be isolated in your mind that you are somehow blessed that no one can possibly match or exceed you method.

    I would venture to say that if you looked there are much more valuable free methods with great documentation abounding all over the internet. So I would humble myself and fully disclose what you have so that at least you can understand you really have nothing at all.

    Do this before you loose any money and you will be wiser for it. mb
     
  9. travis

    travis

    Yes, there are good systems out there, Mark. You are right, but you don't know which ones they are, and that's because no one (on this forum either) ever comes and tells you - this is how my system works and i made this much money with it. So I will behave just like everyone else on this forum.

    Sure you build your system with ideas that already are out there, just like you write a beautiful novel with nothing more than letters from the alphabet. But knowing which are the right things to use makes the whole difference. If you have a system that works, why don't you let everyone have the file? So we can test it, too. Just like you are suggesting me to do. I don't think you are going to do it, and I don't think you find that many people doing it here.
     
  10. ok, what market would you like a system on? pick one. mb

    btw i agree with you generally, but there are some exceptions. myself and some others from the old omega list still give give give even to people we don't know.

    do a search sometime you will change your mind.

    http://www.purebytes.com/archives/omega/
     
    #10     Oct 27, 2007