How To Build an Automated SCT Trading System (Part 2)

Okay, here's a chart. It shows the tapes, double volume bars for each price bar, and it also shows the dom size.

What I want to focus on right now is algorithmically identifying peak volume.


<img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=2109202" height="600" width="800">
Click for larger view


As you can see, all the "turns" happened at peak volume levels.

Any ideas anyone? I'm able to do it as a human pretty well since the volume is rising for a number of bars and then it hits the left line of that tape so I reverse.

Notice that there are caveats because in the first long traverse on the chart, it hit an extreme volume peak, traversed to the other side of the tape and then kept climbing!!

So peak volume doesn't always mean the tape will reverse entirely.

Wayne
 

Attachments

Quote from frostengine:

I have tried automating a strategy similiar to this and found it very dificult making an algorithm that is capable of auto drawing channels...

I am still trying to do it, but what the mind can see very easily its hard to teach the computer how to see it...
More...

Jack posted some interesting diagrams here. The striking thing is how he outlines TL's spawning from the current bar and sweeping back (i.e. Right To Left) as he mentions. The thinking is different from the traditional "Left to Right" mindset and was an AHA for my thinking regarding a design pattern. It may provide some seeds for thought.
 
I ran though this thread, and cant resist posting some comments.

After 4 years in R&D, making my own ATS, complete with charting, back testing, even visual development of my algorithms – back testing 10.000´s of ideas, adding more features, I want to comment:

If you post your ideas in public, they will never be profitable - since, if all uses the same golden trading strategy, it will not be profitable any more. The ATS "jam" is interesting and may give traders some ideas, but if you want to use it as trading for a living, this is a waste of your time - however if this is a hobby, and you dont care if it will make you profits at the end, then its an other story. But the readers should know, that this is a "jam", and noting profitable will be posted, as to the above.

Instead, I would recommend that 2-3 traders gets together, keeping the results in the group.

Further, in order to make a true ATS, you will have to book at least 4 years of your time - full time - to make an ATS which is useable. At start you think that one chart is sufficient, then later you realize that you have to add more charts, then you need to add support for multiple symbols, you add multiple symbols - all the features you think wont be necessary, will be necessary in the ATS. Just to warn you what it takes, and what kind of effort it takes to make - and maintain - a useable ATS. Just ask my whife.

Now 4 years later, its a question whether it was worth the effort. Well, I did finally manage to develop my own trading strategy, but it took 4 years, mainly because I had to continue adding new features to the ATS, needed for the development. Also have a full ATS platform, propably the most advanced around, and made some patents, so both the professional feeling of having created something useful has been met, but also finally to have my own profitable strategies - which I dont share with anybody.

I hope my reply wasnt too negative, but this is my own experience, and perhaps it can help others. I will be happy to give some tips, but not on how to develop trading strategies.

.
 
Quote from ticktrader:
I ran though this thread, and cant resist posting some comments.
After 4 years in R&D, making my own ATS, complete with charting, back testing, even visual development of my algorithms – back testing 10.000´s of ideas, adding more features, I want to comment:
If you post your ideas in public, they will never be profitable - since, if all uses the same golden trading strategy, it will not be profitable any more. The ATS "jam" is interesting and may give traders some ideas, but if you want to use it as trading for a living, this is a waste of your time - however if this is a hobby, and you dont care if it will make you profits at the end, then its an other story. But the readers should know, that this is a "jam", and noting profitable will be posted, as to the above.
Instead, I would recommend that 2-3 traders gets together, keeping the results in the group.
Further, in order to make a true ATS, you will have to book at least 4 years of your time - full time - to make an ATS which is useable. At start you think that one chart is sufficient, then later you realize that you have to add more charts, then you need to add support for multiple symbols, you add multiple symbols - all the features you think wont be necessary, will be necessary in the ATS. Just to warn you what it takes, and what kind of effort it takes to make - and maintain - a useable ATS. Just ask my whife.
Now 4 years later, its a question whether it was worth the effort. Well, I did finally manage to develop my own trading strategy, but it took 4 years, mainly because I had to continue adding new features to the ATS, needed for the development. Also have a full ATS platform, propably the most advanced around, and made some patents, so both the professional feeling of having created something useful has been met, but also finally to have my own profitable strategies - Which I dont share with anybody.
I hope my reply wasnt too negative, but this is my own experience, and perhaps it can help others. I will be happy to give some tips, but not on how to develop trading strategies.
More...
For a second, I thought this post was from our friend ticktrade.

ticktrader: I am glad you have found your way.
I hope our discussion here will in no way diminish your "edge" in the market.

Good trading to you.
 
I completely agree. I hope that I don't mislead anybody into thinking that building an ATS is easy or quick. It takes a lot of long hard work.

Obviously, when it comes to coding and programming, much like any other profession, some are much quicker than others. But it still takes a comparatively gargantuan amount of work to succeed.

I will ADD one more item that ticktrader failed to mention. He referred to all the fruitful successful steps to building an ATS. What really takes the most work is all the things you spend time building that turn out to be WRONG or failure and you have to throw them away, rework them, etc. till you get something right.

It's like Edison with the first light bulb, he failed 1,000 times before succeeding.

Sincerely,
Wayne

Quote from ticktrader:

I ran though this thread, and cant resist posting some comments.

After 4 years in R&D, making my own ATS, complete with charting, back testing, even visual development of my algorithms � back testing 10.000�s of ideas, adding more features, I want to comment:

If you post your ideas in public, they will never be profitable - since, if all uses the same golden trading strategy, it will not be profitable any more. The ATS "jam" is interesting and may give traders some ideas, but if you want to use it as trading for a living, this is a waste of your time - however if this is a hobby, and you dont care if it will make you profits at the end, then its an other story. But the readers should know, that this is a "jam", and noting profitable will be posted, as to the above.

Instead, I would recommend that 2-3 traders gets together, keeping the results in the group.

Further, in order to make a true ATS, you will have to book at least 4 years of your time - full time - to make an ATS which is useable. At start you think that one chart is sufficient, then later you realize that you have to add more charts, then you need to add support for multiple symbols, you add multiple symbols - all the features you think wont be necessary, will be necessary in the ATS. Just to warn you what it takes, and what kind of effort it takes to make - and maintain - a useable ATS. Just ask my whife.

Now 4 years later, its a question whether it was worth the effort. Well, I did finally manage to develop my own trading strategy, but it took 4 years, mainly because I had to continue adding new features to the ATS, needed for the development. Also have a full ATS platform, propably the most advanced around, and made some patents, so both the professional feeling of having created something useful has been met, but also finally to have my own profitable strategies - which I dont share with anybody.

I hope my reply wasnt too negative, but this is my own experience, and perhaps it can help others. I will be happy to give some tips, but not on how to develop trading strategies.

.
More...
 
Something exciting happened last night.

During practicing timing the market with the market replay tool, I realized that I was being inconsistent--successful a while and then I blow it with a blunder.

So I decided to try logging the way Jack and Spyder recommend. Here's what happened.

Most interestingly, after practicing manually writing a bar/bar (and intrabar) log of what's happening with C (continue) X (change) etc. Some almost magical things happened.

1. It brought tremendous clarity to the trades and I was able to pick up much better, precise entries and exits.

2. It virtually eliminated emotions related to trading like excitement when the market is going with you or discouragement when it turns (temporarily) against, etc. The mind was too busy pursuing MADA.

3. Next, I realized that the logging could easily be automated since I was having trouble keeping up with writing at the accelerated rate of 5 minutes bars running for only 1 minute.

4. After automating the logging it became much simpler and more accurate. It's almost uncanny how accurate the logs are which post every 30 seconds (6 seconds on accelerated replay).

Frankly, it now appears possible to take the logging and automate some trade rules because now all I do discretionary is follow what the logging says to do. It makes money. I don't feel I have sufficiently tested this over enough days yet. But so far its better than anything I saw before.

If you want details about logging, please search the forums, there's already tons of material on that issue. Jack and Spyder recommend coming up with your own comfortable format and notation after some suggestions they give.

Wow. It seems to work.

Wayne
 
Quote from greaterreturn:

..
2. It virtually eliminated emotions related to trading like exciting when the market is going with you or discouragement when it turns (temporarily) the wrong way, etc. The mind was too busy keeping pursuing MADA.
..
Wow. It works.

Wayne
More...
especially true for manual trader.
 
Quote from nkhoi:

especially true for manual trader.
More...


Agreed, I can see now that this was the one missing element when I attempted day trading discretionary years ago.

It was a different strategy than SCT. But I did the other stuff Jack recommends like housekeeping and debriefing each day.

But MADA is really a boost to effectiveness and consistency. It also makes it more relaxing.

One more thing. It was the first time I took my eye off the amount of money I was making or losing. In fact I realized after about 5 trades that I never thought of checking the report to see how many points I was making.

It felt irrelevant. The trades were working, I was busy making decisions per bar.

It was very cool. I'm convinced I could do this discretionary and will probably try it with real money next week.

But I'm intrigued now with trying to automate the decision and action step of the MADA logging now. Then it could become completely hands-free or at least get closer yet.

Wayne
 
Quote from greaterreturn:
Okay, here's a chart. It shows the tapes, double volume bars for each price bar, and it also shows the dom size.
What I want to focus on right now is algorithmically identifying peak volume.

<img src="http://www.elitetrader.com/vb/attachment.php?s=&postid=2109202" height="600" width="800">
Click for larger view

As you can see, all the "turns" happened at peak volume levels.
Any ideas anyone? I'm able to do it as a human pretty well since the volume is rising for a number of bars and then it hits the left line of that tape so I reverse.
Notice that there are caveats because in the first long traverse on the chart, it hit an extreme volume peak, traversed to the other side of the tape and then kept climbing!!
So peak volume doesn't always mean the tape will reverse entirely.
Wayne
More...

For those who are interested, I have posted the code for building tapes in EasyLanguage here:
http://elitetrader.com/vb/showthread.php?s=&postid=2109583#post2109583
 
Quote from Tums:

For those who are interested, I have posted the code for building tapes in EasyLanguage here:
http://elitetrader.com/vb/showthread.php?s=&postid=2109583#post2109583
More...

Tums, they look cool!

When I read about tapes in "Channels to Building Wealth", it seems Jack extends them past the first 2 bars.

Yours reset on every single bar.

Was that an intentional decision or improvement?

Admittedly, it's much more complex to extend the tapes out for succeeding bars, I think.

Or am I misunderstanding the term "tape". does tape only refer to the initial 2 bar patterns. If so, then what I have in the chart above maybe are called ST (short term) channels.

Please clarify, Tums, or someone.

Sincerely,
Wayne
 
You must log in or register to reply here.
Back
Top