I am presenting here something I noticed in hopes that it will be useful and maybe we can build a system or something from this. I hope for feedback from more experienced members but any constructive comment from everyone is really appreciated.
I really like the Donchian Channels. They can signal breakouts that can enable us to catch major swings (I don't think in forex we can catch the whole trend). Unfortunately they fail most of the time, especially in current market conditions. They break, take us in just to drag us back in the red and eventually in a losing trade. But what if, when it fails, we take that failure and trade it?
In the following chart I have drawn vertical lines where I consider there would be good candidates to take entries.
The rules are simple. We need a breach of the 20 days Donchian Channel and a close behind the break point accompanied with a negative ADX(14) reading.
Some points to consider
* Regarding the ADX(14) it's just the default setting
* The idea is that the price comes from the other side with momentum so the reading will be up. Once reached there the momentum will fade so the reading will turn negative. If it doesn't turn negative then maybe it has some more steam to go.
* It doesn't matter if until yesterday ADX was positive. If today is lower than yesterday that is good enough
* Donchian 20 is well know and more prone to whipsaws than higher number of days. That's what we are trying to take advantage.
* A breakpoint is considered valid if the line has remained flat for at least five days.
* In strong trend we need a double top/bottom kind of formation to consider the signal
* We must enter with stop orders (like we trade the breakouts).
Unfortunately there are still points I haven't figured out. I am attaching another chart. This time EURUSD. With tick are marked the valid entries and with X invalid entries just to demonstrate my points above.
Unfortunately there are entries that would have been taken and would have ended as whipsaws. It is true that I haven't tested this thoroughly and this is the chart that displays the most failures but how can we filter this further? Also where should we get out? One idea that comes to mind is not to move the stop to break even but once it reaches a point of interest (like for example the other side of the Channel) and we get a signal for the opposite direction, then we move our stop behind that signal bar. This seems reasonable but I would like to hear other ideas and improvements.
I really like the Donchian Channels. They can signal breakouts that can enable us to catch major swings (I don't think in forex we can catch the whole trend). Unfortunately they fail most of the time, especially in current market conditions. They break, take us in just to drag us back in the red and eventually in a losing trade. But what if, when it fails, we take that failure and trade it?
In the following chart I have drawn vertical lines where I consider there would be good candidates to take entries.
The rules are simple. We need a breach of the 20 days Donchian Channel and a close behind the break point accompanied with a negative ADX(14) reading.
Some points to consider
* Regarding the ADX(14) it's just the default setting
* The idea is that the price comes from the other side with momentum so the reading will be up. Once reached there the momentum will fade so the reading will turn negative. If it doesn't turn negative then maybe it has some more steam to go.
* It doesn't matter if until yesterday ADX was positive. If today is lower than yesterday that is good enough
* Donchian 20 is well know and more prone to whipsaws than higher number of days. That's what we are trying to take advantage.
* A breakpoint is considered valid if the line has remained flat for at least five days.
* In strong trend we need a double top/bottom kind of formation to consider the signal
* We must enter with stop orders (like we trade the breakouts).
Unfortunately there are still points I haven't figured out. I am attaching another chart. This time EURUSD. With tick are marked the valid entries and with X invalid entries just to demonstrate my points above.
Unfortunately there are entries that would have been taken and would have ended as whipsaws. It is true that I haven't tested this thoroughly and this is the chart that displays the most failures but how can we filter this further? Also where should we get out? One idea that comes to mind is not to move the stop to break even but once it reaches a point of interest (like for example the other side of the Channel) and we get a signal for the opposite direction, then we move our stop behind that signal bar. This seems reasonable but I would like to hear other ideas and improvements.