Tools For Trading using the Price Volume Relationship
http://www.elitetrader.com/vb/showthread.php?s=&postid=2054295#post2054295
http://www.elitetrader.com/vb/showthread.php?s=&postid=2054295#post2054295
Price action and market participant
- Thread starter phattails
- Start date
Trader666,
Hmm. Did you miss what I said? I haven't "profitably" coded SCT yet. I said what I have coded so far looks good but it still loses pips over all until I refine some of the rules.
But I have built a superior platform. But neither you or anyone else would consider it superior because it's custom built to my prefereneces.
You're obviously very smart so you realize, of course, that everything that anyone says about anything is relative.
So my platform lacks a plethora of features others might consider necessary. But it has an abundance of features absolutely essential to me.
By the way, I strongly recommend staying away from the SCT and PVT type strategy unless someone has a very detail oriented personally.
It seems SCT and PVT appeal to detail oriented people like myself and was created by those. Birds of a feather.
Our ilk often drive other "big picture", K.I.S.S. type of people out of their minds. You can spot the difference a mile away by the length of people's posts.
Probably, if you read this far in the post already, you might have detail oriented tendencies.
That's because us detail oriented people often go to lengths that others cringe to even consider. Worse yet, we sometimes tend to push others in that direction.
Just to give you an idea, I spent thousands of hours studying strategies over 14 years. I traded discretionary using options, futures, and also day traded as part of my self training. I have experimented with every type of platform or technology from all the way from TradeStation to NeuroNets. Studied money management, etc.
You mention that I built this in a few months. Largely that's true but I had code from stuff I built while trying to use commercial systems, like algorithms, time based tick filter, etc.
To build this current platform it has cost many hundreds of hours over weekends, evening, and entire nights.
So I don't know for certain that I can figure the rest of SCT and PVT out...but I think so.
What got me going in spite of all the negative hype was something I learned about the DOM from posts by spyder and Jack which greatly reduced the risk of another ATS I already had.
That, needless to say, impressed me that they at least have something right. The DOM was one area I had never studied or understood before.
After that, as a highly detail oriented person, I am drawn to the observation that my other contrarian ATS was out of the market the majority of the time and leaving tons of money on the table.
To give you a "break", since my detail oriented, overly hard working nature bothers you, I suggest being kind enough to put me on ignore to avoid enduring my posts at all. I will return the favor right now to avoid writing long detailed (time consuming) replies to you.
Sincerely,
Wayne
Tums,
Thanks for this link. I'll check it out. So much to learn.
Wayne
Thanks for this link. I'll check it out. So much to learn.
Wayne
So far so good. I got the PV&S relationship working.
But, it's still not profitable at the moment.
I can see why. It's because I was trying it without channels.
So it does extremely well at catching all the significant moves by entering right at the low and exiting at the high.
However, it gets simply PV relationship gets killed in the sideways or choppy areas without the use of channels.
After implementing the channels, that will make those areas profitable TOO! That's because it can buy and sell each traverse of the channel.
Then it only trades the bigger moves on BO from the channel with sufficient volume to justify.
I'm having one of Jack's AH AH! moments.
I already have LR code and code for drawing lines and calculations from another ATS so it won't take long to incorporate that into this ATS.
Wayne
But, it's still not profitable at the moment.
I can see why. It's because I was trying it without channels.
So it does extremely well at catching all the significant moves by entering right at the low and exiting at the high.
However, it gets simply PV relationship gets killed in the sideways or choppy areas without the use of channels.
After implementing the channels, that will make those areas profitable TOO! That's because it can buy and sell each traverse of the channel.
Then it only trades the bigger moves on BO from the channel with sufficient volume to justify.
I'm having one of Jack's AH AH! moments.
I already have LR code and code for drawing lines and calculations from another ATS so it won't take long to incorporate that into this ATS.
Wayne
Quote from rawfist:
What kind of a moron buys on a trendline break?
The brown channel is a traverse. It ENDED at that bar, hence it did a breakout. However, you need to realise a traverse is just 1/2 of a channel, so when you get this breakout (which failed to provide a counter traverse) we expect a similar traverse.
I'm pretty sure he deliberately shows examples right on the edge (expert if you will) to fool around with sceptics like you.
It's Ironic because you're so convinced that you just don't get it he's fooling you. so right back at you, moron
(ET keeps sending me emails for posts by people on ignore. Oh bother.)
Anyway, about price action:
It created a major AH HA regarding price action yesterday while learning FTT (Failure to Traverse).
Background first, it was already clear t me in PA (Price Action) that prices, S/R, and channels operate diagonally rather than horizontally . That goes against CW (conventional wisdom).
FTT was a major AH HA! because an FTT is, in CW PA parlance, a higher high (HH) or lowerlow (LL). However, FTT recognizes that a HH or LL failed to traverse as far DIAGONALLY as the last HH or LL. Therefore, the HH or LL in that instance indicates an end to the trend whereas, CW S/R feels that HH is already bullish and LL is always bearish.
That fact can really bite if you consider a HH (which is an FTT) bullish and enter on a pull back only to see the market continue dropping to a new LL.
It takes a paradigm shift to read price action diagonally rather than horizontally.
If you study FTT's on a chart, it makes a WORLD of difference in correctly picking off highs and lows along with other more ordinary bounces off the diagonal S/R lines.
Now, automating an FTT is an interesting and tricky problem. But I'll have it working soon.
The other entries and exits for bounces off the diagonal S/Rs are working as of yesterday in my ATS.
It really is very beautiful thing to see all those reversals at exactly the highs and lows.
Adding FTT will fix most of the problem of predicting BO (break out) in advances--but not all the time. It seems that's were the DOM and volume action come into play.
I've got to study all the channel concepts again to see if there's any other necessary tidbits.
Sincerely,
Wayne
Anyway, about price action:
It created a major AH HA regarding price action yesterday while learning FTT (Failure to Traverse).
Background first, it was already clear t me in PA (Price Action) that prices, S/R, and channels operate diagonally rather than horizontally . That goes against CW (conventional wisdom).
FTT was a major AH HA! because an FTT is, in CW PA parlance, a higher high (HH) or lowerlow (LL). However, FTT recognizes that a HH or LL failed to traverse as far DIAGONALLY as the last HH or LL. Therefore, the HH or LL in that instance indicates an end to the trend whereas, CW S/R feels that HH is already bullish and LL is always bearish.
That fact can really bite if you consider a HH (which is an FTT) bullish and enter on a pull back only to see the market continue dropping to a new LL.
It takes a paradigm shift to read price action diagonally rather than horizontally.
If you study FTT's on a chart, it makes a WORLD of difference in correctly picking off highs and lows along with other more ordinary bounces off the diagonal S/R lines.
Now, automating an FTT is an interesting and tricky problem. But I'll have it working soon.
The other entries and exits for bounces off the diagonal S/Rs are working as of yesterday in my ATS.
It really is very beautiful thing to see all those reversals at exactly the highs and lows.
Adding FTT will fix most of the problem of predicting BO (break out) in advances--but not all the time. It seems that's were the DOM and volume action come into play.
I've got to study all the channel concepts again to see if there's any other necessary tidbits.
Sincerely,
Wayne
FYI, about that NTES diagram posted earlier. I studied it. It only shows the major channels. Experts will look at channels on multiple fractals.
So the NTES appeared to be a breakout on the major channel when the chart stopped.
However, on the secondary chart, things developed differently.
The market did BO to the upside of the major chart. But anybody good at SCT would have additional channels either on their chart or in their head at a smaller time frame.
So it's obvious that around 23.08 or so, the market broke out of the upside short term channel to the down side.
Anyway, who would have kept holding on that stock while it dropped like a stone? That's silly, you continue to adapt to the changing landscape.
You can never "predict" what happens. You can anticipate and be ready when it does, however.
Anyway, adding another person to ignore due to insulting language.
Take care.
Wayne
So the NTES appeared to be a breakout on the major channel when the chart stopped.
However, on the secondary chart, things developed differently.
The market did BO to the upside of the major chart. But anybody good at SCT would have additional channels either on their chart or in their head at a smaller time frame.
So it's obvious that around 23.08 or so, the market broke out of the upside short term channel to the down side.
Anyway, who would have kept holding on that stock while it dropped like a stone? That's silly, you continue to adapt to the changing landscape.
You can never "predict" what happens. You can anticipate and be ready when it does, however.
Anyway, adding another person to ignore due to insulting language.
Take care.
Wayne
Hi Wayne,
I too have been focused on coding up the modules which Jack has outlined in other posts. Particularly, the ones where he builds up each module taking inputs and producing outputs which become storage for analysis and decisions or inputs into subsequent modules. The application is very much in the "prototype" phase of development. But since you were speaking of development of the PE principles, I thought I would share.
I have just begun the analysis into the channel annotation module (3 levels of channels). I am working on some prototype code for that using the P/V relationship and channel rules and hope to begin incorporating these into the program soon. I am amazed how brilliant the mind is and well... unfortunately, I am just not that great of a developer, but I will press forward with the development of the application as it has been very enjoyable and enlightening.
For now, I have been working on coding the cases, volume, and P/V modules and performing analysis and decision making on these output data points. I have attached an image of todays results from the existing prototype code and here is a link to the replay from today (ES 5 minute). It is an AVI file (and pretty big too ~ like 50 mb, so it takes some time to buffer). I am pleased with the trending results too. But there is definately room for refinement and the channels will add a much needed "container" to help make more informed decisions and more accurate trades, IMHO. The prototype has some quick code in place to begin evaluating the analysis and decision making logic. For instance, it always enters on Bar 3 and always exits on bar 78 (on a 5 minute chart). Every trade in between is a reversal and every trade is made at the close of the bar. I have no medium or fine level indicators coded at this point (i.e. YM, STR/SQU, DOM, TICK, etc...).
The next keys I am focused on (besides the channels module) are to refine the logic for filters and context, as they relate to analysis and decision making (i.e what to do in internals, Peak Volume, etc..., when to take the signal for change and when to filter for context and so on). This prototype is a Windows Forms application, developed using C# .NET 2.0... It is very crude at this point but is offering me a great bed for developing and testing strategies using Jacks principles.
Anyways, just thought I would share and wish you all the best on your development.
I too have been focused on coding up the modules which Jack has outlined in other posts. Particularly, the ones where he builds up each module taking inputs and producing outputs which become storage for analysis and decisions or inputs into subsequent modules. The application is very much in the "prototype" phase of development. But since you were speaking of development of the PE principles, I thought I would share.
I have just begun the analysis into the channel annotation module (3 levels of channels). I am working on some prototype code for that using the P/V relationship and channel rules and hope to begin incorporating these into the program soon. I am amazed how brilliant the mind is and well... unfortunately, I am just not that great of a developer, but I will press forward with the development of the application as it has been very enjoyable and enlightening.
For now, I have been working on coding the cases, volume, and P/V modules and performing analysis and decision making on these output data points. I have attached an image of todays results from the existing prototype code and here is a link to the replay from today (ES 5 minute). It is an AVI file (and pretty big too ~ like 50 mb, so it takes some time to buffer). I am pleased with the trending results too. But there is definately room for refinement and the channels will add a much needed "container" to help make more informed decisions and more accurate trades, IMHO. The prototype has some quick code in place to begin evaluating the analysis and decision making logic. For instance, it always enters on Bar 3 and always exits on bar 78 (on a 5 minute chart). Every trade in between is a reversal and every trade is made at the close of the bar. I have no medium or fine level indicators coded at this point (i.e. YM, STR/SQU, DOM, TICK, etc...).
The next keys I am focused on (besides the channels module) are to refine the logic for filters and context, as they relate to analysis and decision making (i.e what to do in internals, Peak Volume, etc..., when to take the signal for change and when to filter for context and so on). This prototype is a Windows Forms application, developed using C# .NET 2.0... It is very crude at this point but is offering me a great bed for developing and testing strategies using Jacks principles.
Anyways, just thought I would share and wish you all the best on your development.
Quote from ehorn:
Hi Wayne,
I too have been focused on coding up the modules which Jack has outlined in other posts. Particularly, the ones where he builds up each module taking inputs and producing outputs which become storage for analysis and decisions or inputs into subsequent modules. The application is very much in the "prototype" phase of development. But since you were speaking of development of the PE principles, I thought I would share.
I have just begun the analysis into the channel annotation module (3 levels of channels). I am working on some prototype code for that using the P/V relationship and channel rules and hope to begin incorporating these into the program soon. I am amazed how brilliant the mind is and well... unfortunately, I am just not that great of a developer, but I will press forward with the development of the application as it has been very enjoyable and enlightening.
For now, I have been working on coding the cases, volume, and P/V modules and performing analysis and decision making on these output data points. I have attached an image of todays results from the existing prototype code and here is a link to the replay from today (ES 5 minute). It is an AVI file (and pretty big too ~ like 50 mb, so it takes some time to buffer). I am pleased with the trending results too. But there is definately room for refinement and the channels will add a much needed "container" to help make more informed decisions and more accurate trades, IMHO. The prototype has some quick code in place to begin evaluating the analysis and decision making logic. For instance, it always enters on Bar 3 and always exits on bar 78 (on a 5 minute chart). Every trade in between is a reversal and every trade is made at the close of the bar. I have no medium or fine level indicators coded at this point (i.e. YM, STR/SQU, DOM, TICK, etc...).
The next keys I am focused on (besides the channels module) are to refine the logic for filters and context, as they relate to analysis and decision making (i.e what to do in internals, Peak Volume, etc..., when to take the signal for change and when to filter for context and so on). This prototype is a Windows Forms application, developed using C# .NET 2.0... It is very crude at this point but is offering me a great bed for developing and testing strategies using Jacks principles.
Anyways, just thought I would share and wish you all the best on your development.
Good start. As your channels module comes up to speed you will drop out the fractal shift that is going on here and there in your attachment. I couldn't download the link because this computer has high security level.
Also as you do the coding on the internals you will take out the OB inconsistency that is showing. Notice on the RS before the first of two OB's you rode through the OB. Contrarily, on the next OB you did the first of two reversals where a hold through on RS would have made more coarse level profits and concurrently eliminated any possible losses other than the "coarse level" reversals to stay on the right side of the market.
In a while you will see the value of carrying forward status stuff for several bars. This will get in the picture as you adjust your thinking to just how to keep contemporary data available and eliminate history that is no longer in play. This is a control mechanism for establishing the variable width of the non stationary window. The look back is done with a landmark which is a diagonal output of one of the daigonal boxes.
I see by your link that you make it possible to go somewhere and pick up something that is beyond ET. Do you know how I could put up, say 50 pages of text or 20 illustrations referenced into such text?
I feel I can cut through some of the incessant stupidity the disparagers continue to misunderstand if I could post a bunch of final statements on various applications of the pool extraction algorithm.
I'll post some of the recent comments I have deleted here as well. They will be of interest to both of you.