a trading strategy based on Elliott Wave Theory - page 149

 
<br / translate="no"> There really is no such postulate. "The channel is broken up" is a fact, "price IS going up" is also a fact for the moment, but no one is saying anything about what happens next. The parameters of a new channel change as it develops, the direction may change and then a new channel may start. However, the picture will explain it to you.


Apparently, I did not quite get your idea right earlier :o)))
 
The picture in the first post of this page reminded me of the picture my friend from America, who plays stock indices for a living, is looking at. He used to play Forex, but he says that it is more convenient to earn on indices.
At home, he has set up a professional market place with 7 monitors and 2 computers. For about 3 years now he has been in this business. He tried a lot of different things, and the last thing he stopped at was just channels, which he draws approximately as described above (a new channel at the breakdown of the current one). He's been drawing them for probably over a year now. I asked him how he got to this point. He says it's just intuitive, he looks in the monitor 24 hours a day. He didn't use mathematics, he says he has a very complicated relationship with it. He's still a long way from being a millionaire, but he makes enough to live on. He manages to earn even more than his mates in their regular jobs.
 
The picture in the first post of this page reminded me of the picture my American acquaintance, who plays stock indices for a living, is looking at. <br / translate="no"> ...
He's still a long way from being a millionaire, but he earns enough to live on. He manages to earn even more than his buddies in their regular jobs.

That's right. That's it. The idea lies on the surface, but it was difficult to realize it algorithmically, there were too many nuances. It is not so much a computational process as it is image recognition. It is not so easy to do by simple means. :-)
But there is a positive moment - it means that you can earn "even more" with this method. :-))
 
The idea lies on the surface, but it was difficult to implement it algorithmically, there were too many nuances.

I understand that the indicator that implements the above described rules is ready, then we are looking forward to testing the strategy in the tester. At least preliminary.
 
Yuri, a quick question. Are the constructed channels in the picture the result of recalculation "in the past" or does your system build such channels in "real time", i.e. we can say - without errors. There are breakdowns on the chart and logically (of course, in my understanding of your approach), the channel should change, i.e. rebuild, and as a result (of course, I could be easily mistaken), have a slightly different appearance (but this is, as they say, "by eye").
 
2 solandr
I take it that the indicator that implements the rules described above is ready, then look forward to testing the strategy in the tester.

I wrote that there are still some refinements to be made. In particular, there are two situations in the price behavior that I would like them to handle differently. When it is done, we can say that the indicator is ready. In fact, it cannot be called an indicator. It may be called a chart indicator.

As for the strategy, I think that just channels alone are not enough to build it. I believe the strategy should have 4 analytical components and channels should be only one of them. In addition, in order to be able to use them a statistical analysis of some parameters should be performed. The algorithm is formulated by me and therefore carries the imprint of subjectivity. Statistics should show in which situations and to what extent this algorithm and the market understand each other. Then it will become clearer when and what decisions can be made on these channels.

2 grasn
The constructed channels in the picture are the result of recalculation "in the past" or your system builds such channels in "real time", i.e. we can say - infallibly. There are breakdowns in the chart and logically (of course, according to my understanding of your approach), the channel should change, i.e. be rearranged, and as a result (of course, I can easily be mistaken), it should look somewhat different (but this is, as they say, "by eye").

I gave up recalculations into the past right away. The channels are calculated only into the future. The top from which the channel is counted does not change. The channel itself, of course, may change. If this change is not a breakout, it can even turn in the opposite direction. But if it is a breakout, a new channel starts to be built. The problem is to distinguish the evolution of an old channel from its destruction into a new one. So you are absolutely right "by eye". :-)
 
<br / translate="no"> I gave up recalculation to the past right away. Channels are only calculated into the future. The vertex from which the channel is calculated does not change. The channel itself may change, of course. If this change is not a breakout, it can even turn in the opposite direction. But if it is a breakout, a new channel starts to be built. The problem is to distinguish the evolution of an old channel from its destruction into a new one. So you are absolutely right "by eye". :-)


Yuri, thanks for the answers. I was just confused by the pronounced breakdowns, especially on the blue channel, left unnoticed by the algorithm. According to my understanding, based only on the stated logic corrected by my cleverness (:o), the channel should realign (not only the blues channel), and the search for appropriate extrema should start from the breakdown point according to the rules and involve extrema that are obvious (to the eye, of course) for the new channel.

Hence, at the moment (when the zero bar was "at the breakdown point" at the time),
- either the overshoot criterion 2SCO did not work (not in the sense of a programming error, but simply there was no overshoot signal, i.e. the 2SCO limits were in other ranges),
- or maybe the controlling component of your system decided based on additional criteria, that it is too premature to rebuild the channel because of the bar that "broke through" 2SCO and should go on fluctuating with the batch. :о)))

PS: Sorry for the intrusiveness, It's all out of curiosity. :о)
 
grasn
You are absolutely right. What you see in the picture is, of course, a fixed situation, i.e. its last state left in the past. But the current channel lives and changes. And, accordingly, the algorithm that builds and maintains those channels lives and reacts to those changes.
 
Yuri, am I correct in assuming that the very value of the RMS during the development of the channel from its fixing is recalculated with each new bar?
 
Vladislav, on the picture represents the indicator you wrote, which was based on Auto Harmonic Patterns indicator (Version 1.5 30/04/2006)? I found only the ex4 file of this indicator on the web, but it doesn't show me anything except its name on the chart. Probably it's some trial version that works until a certain date? Could you share this Auto Harmonic Patterns indicator, if it is possible of course? Or if it's not possible, could you provide a link to purchase it? Thank you!