Does anyone know of Clayburg's Breakout System with Parallel Function?
I am studying Clayburg's parallel function self-adapting systems. It is a very interesting concept! He posted some examples from Omega/TradeStation presentations from '99 and '05. I pasted them into the PowerEditor and got the '99 version to plot as an indicator. The indicator values do not match the system equity curve, as they should. The '05 version will not verify as it is calling an undefined user function called "DCount" that I do not have. I guess it is a day counter but am not sure. Can anyone help?
I read his material on the parallel function and did code up an example. It was a real pain because functions to not have the same abilities as strategies so you must write code to handle such issues. The parallel function would never completely mimic a real trading system. You could get close, but not perfect. In short, my study was dynamically changing the input parameters on his value system based upon which parallel function was performing better. My summery, there was no improvement in trading performance.
The ideas of using strategies as indicators is not new and may be better implemented with TradeStations 9.0 feature of sending data to a spreadsheet. In this way you can have several windows open running the system strategy code with different input values. Their trades can then be sent to a spreadsheet. A "live" system reads the spreadsheet then picks the best system and runs with it.
Seems like it would work, but I have yet to test it.
The following user says Thank You to Jeff65 for this post:
Thanks for your reply. I was beginning to wonder if anyone else out there is stubborn enough to try coding this technique! I have spend about three days working on it and finally got a "parallel function" to output the exact same values as the strategy (caveat, not with LIBBT). The goal is now to get the "parallel function" to dynamically update one if the system variables via looping with arrays and sorting. It is a lot of work and I hope that it pays some dividend besides experience.
I didn't know that TS9 could export spreadsheets. That was an old feature that was lost years ago. Many people have complained. I guess enough to get the feature back. Thanks for the info!
Yeah, TS 9.0 has the ability to read and write to spreadsheets. It also has the ability to pass values between charts - something that was once only available via third-partty plugin (ADE). I've not played with these features very much so I'm not familiar with their full abilities or limitations.
Thanks for the compliment. One of the secrets to getting the parallel function to mimic the system exactly is to force them both to handle minimum fluctuations in the Data1 exactly the same way. TS "knows" that a system cannot order 0.10 points worth of ES but can for SP. The parallel function "knows nothing" about the specifications of the contract. The sys dev has to code the logic in to handle this problem otherwise the results of the two will rarely match.
The TS9 concept of objects is wonderful and it is about time that they (TS) catch up with what could be done in modern OOP languages for decades. Now, if we can only get them to release a production version that is 64 bit and multi-threaded!
I am a 45 year old male who had been trading futures for a few years now.
I have tried some of John Clayburg's systems. I was convinced he was being honest about the results he posted.
After spending a lot of time and money I found a few major flaws in his strategies. These flaws were enough to create losses every single day with his automated systems...although on his site he posts gains. That is false advertising and should not be allowed. I studied his systems a lot and did everything to make his strategies work.....but the losses were consistent and I saw very few gains.
In addition, he is very rude, not replying to emails or returning phone calls. When I finally did get a hold of him to ask him some questions he made me feel like I was bothering him and gave me a lot of one word answers.
He calls himself a doctor yet in reality he is a vet. That is a warning sign right there.
His system costs over $6000. Don't be taken by him. I learned my lesson the hard way. Save your money. I hope I have helped some people.