fxDreema

    • Register
    • Login
    • Search
    • Back to the main page
    • Categories
    • Recent
    • Tags
    • Popular
    • Search

    Open discussion on defining end of trend

    General Discussions
    2
    2
    330
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • O
      osryan last edited by

      Ok all, I imagine most of us grapple with a similar issue when building our algos: keeping it out of market when conditions aren’t right. Also for most of us, I suspect the ideal condition is trend, range is bad but end of trend is devastating. You may not make any money in ranges, but they can be survivable. Ends of trends are violent though and by the time your algo recognizes the trend change, it’s already been crushed numerous times and so quickly that a trailing or BE stop never had a chance to limit the damage.

      So without giving away our secrets on the algos we’ve worked so hard on, anyone have thoughts on dealing with this? For example, I’m looking at two ways right now that I’m not sure yet how to code. First is to have some sort of distance rule so when price gets too far from my MA, trades don’t open. The thinking is when the trend has that blow off top before the end, it gets real far from the MA and the danger of a fast reversion to the mean increases.

      The second idea is some sort of way to define when price leaves an area and enters a new equilibrium above recent high or below recent low. For example, in Renko, this could be defined as the price where brick color change happens. Since price has to move 1x to form a new brick in direction of trend but 2x in the opposite direction, a color change signifies a substantial price movement. So when price breaks that reversal level, I can define that price is indeed corresponding with the trend my indicators are showing. If it doesn’t, then trades halt.

      Anyone have any other thoughts? I personally like the first idea better than the second because it seems simpler but one or both could be missing the mark completely and I won’t know until I code it. Of course for every rule that goes in, the downside is it will filter out winning trades too.

      1 Reply Last reply Reply Quote 0
      • L
        len 1 last edited by

        It sounds like you got some secrets that you respectfully you opt others not to share. This whole forum is on making each other better.

        1 Reply Last reply Reply Quote 1
        • 1 / 1
        • First post
          Last post

        Online Users

        R
        F
        S
        D
        V
        O

        15
        Online

        146.7k
        Users

        22.4k
        Topics

        122.6k
        Posts

        Powered by NodeBB Forums | Contributors