TRO PMSM TREND

[peter_77_peter]

TRO:

Thank you for the speed and the quality of your response.

K.I.S.S. was my intention, TMI my enemy, I´ve just lost my focus!

I think you´ve gotten me completely right, and you got me, i simply wanted too much.

So I´ve brought my focus back to simplicity, banned trend strength, got back to displaying all directional movement as green and red dot in my the PMSM Version of my setup, adding nothing more than blue dots for visually filtering possibly weak trends/range. Trend strength can be interpretated as the validity of directional movement, and this can be read of directly from the patterns of the PMSM, in terms of "the agreements between the red and the green dots".

Thank you for your inspiration, keep up the good work!
best regards from spain,
peter

[TheRumpledOne]

Peter:

You're welcome.

Please post screenshot examples of how you use PMSM.

Thanks.

[peter_77_peter]

TRO:
When I´ve "finishes my homework" I will post my screen, this will be within the next days. The company I am working for insists on keeping some information about the inputs used "blackened". But I guess even without all information shown I´ll still be able to explain some thoughts on what I get from PSMSM Interfaces.

[TheRumpledOne]

So the company uses other people's work for free but doesn't want to share?

[peter_77_peter]

- I´m not the company, I´m just an employee.
- It might be possible that you´ve got me wrong. I just said:
- I want to finish my homework before I show screenshots, because I think my Interface will get better within the next days, lets say until the weekend or the beginning of the new week.
- I definitely will share screenshots, because life gets better (and lovelier) troughout give and take.
- I just have to respect some guidelines of the company I´m working for and therefore have to be creative with sharing the underlying ideas/thoughts of my work without telling every detail of the signals we use within this interface. This does not prevent me from sharing the ideas behind or lets say coding experience with tradestation within this forum. And this is exactly what I´ll do. And I guess that is in some way even more contributive, because you kind of get distracted from the big picture/ideas behind the setup if you look at all it´s details. Probably everybody has (should have) different setup preferences about what i.e. signals feels comfortable. So this should not be about copying other peoples indicators/setup 1:1 using them as a black box and don´t know how or why. But it should be about getting inspired on the idea behind, and sharing the know how how to adapt these ideas within the own setup.
- I think this is legitimately, and likewise giving away "the big picture trading strategy" ebook for free, and selling the "breakout bounce trading system" for 500 cash. I´m looking forward reading the big picture trading strategy, and if buying "breakout bounce" seems reasonable somewhen, I´d then enjoy paying for your work that I highly appreciate.

[TheRumpledOne]

LOL... I didn't say it was you... I know it's the company.

[albeit01]

thank you TRO for pmsm_trend

am very excited about using it

has already saved me from entering bad calls

thanks again

albeit01

[RuGO]

Avery, I took the liberty of tweaking the code so as to incorporate the button into the matrix. Changed the contrast, so I can superimpose just one more indicator (lol, as we cannot seem to get enough).

The one question I have for th users is: has anyone found any specific indicator settings that increases the efficiency of this system? I don't think writing a script for backtesting and optimizing is in my schedule anytime soon



Cheers,

Russ

[peter_77_peter]

RuGO (and all others interested in discussing usage / effiecency of TRO_PMSM_TREND):

About your question concerning efficiency of TRO_PMSM_TREND:

I really like TRO PMSM Trend because its visual interface fits the trading style I`m working on. My clues that I´ve found out when I played with it on stocks a lot was:

- the 5 / 10 / 20 / 30 length setup for the underlying XMAs seems well chosen
- Dislike 1: It does not differentiate between strong and weak trend / increase/decrease/slope of the underlying XMAS.
- Dislike 2: It does not recognize when the movements in the Stock are barely trendless and therefore generates inefficient signals in "times of ranges/sideways movements".

- So what I did about 1 is: I`ve had a look at the slope of the XMAs at a percentual basis and found some triggers/thresholds that seemed stable to discrete "strong" and "weak" ups/downs. These triggers are dependend on the timebase you look upon, but theres some beauty in the numbers, and the triggers/thresholds can be calculated in some factor in dependency of the timebase. This seems plausible, because the amplitude just falls of with some factor if you "zoom in" the timeframe. If the thresholds fullfill I´ve set the plotcolor to "darkcolor", therefore strong upmove "darkgreen", strong downmove "darkred", weak upmove "lightgreen" (normal green), weak down move "lightred" (normal red). Basically that is some kind of discretization of the sinal levels to squeeze out some information, can be called filtering as well, with comes with some advantages and some disadvantages, but for me it seems to fit the problem I´ve had with dislike 1 of TRO_PMSM_Trend.

- What I did about 2 is: I´ve adapted the principle of chandes RAVI to the TRO PMSM TREND. Therefore I´ve been looking at the absolute difference for each underlying XMA with a XMA "shorter in length by a factor of ten". Therefore I´ve leveled out "weak range" and "strong range" signals for each signal dot, though I´ve added 4 more dots, each one colored gray if there is Trendiness, lightblue for "weak trend dynamics/weak range", darkblue for "no trend dynamics/strong range". That is basically Chandes Ravi principle adapted to the XMAS of TRO_PMSM_Trend. That is basically just a Trendfilter getting very close to the view of each Trend/Dot. It seems to help a lot in filtering out inefficient signals, that is: "all greens" "all reds" in times where trend dynamics is low / range. I´ve found some beauty in numbers with the thresholds again, therefore the thresholds are dependend of the timeframe you´re looking at, but I´ve found some way to calculate it very accurately.

- I explicitly say that this seems to be an helpful addition for the way I`m using TRO_PMSM_Trend. It seems to increase the efficiency for me very solid. But in adding some complexity or filtering theres allways a tradeoff, I find my extensions reasonable and sound, but others may say I just complicate things and my additions take away some of the key approach of TRO_PMSM_Trend: K.I.S.S.

- I`ve promised TRO some weeks ago to post screens, and I definitely will. But it will take some more weeks, maybe two. As skateboarders say: "Im in the middle of a trick", and I don´t want to post anything that is not completely done, well tested, really finished. So I just post some thoughts for now.

I´m very interested in other opinions of how other users use this indicator, or the pmsm principle on stocks. for me its just one component of a trendfilter, that is watching for trend direction / trend strength / range on a daily basis, with entry triggers of trend following strategies trading in the direction of trend on a 60 min chart / doing nothing if trend dynamic is down / range.

I´m very new to the game, and have no good know-how on oscillators or other things to look at for strategies that are efficient in times when trend dynamic gets down / range. So if anybody has good links, good experiences with doing something good in times when trend following does not. Then that is very interesting for me too.

Besides that, has anybody made good experiences with Portfolio Analyzing tools for Tradestation?

all the best,
Peter

[RuGO]

Peter,

Thank you for the response and for sharing! I agree with you that from a visual perspective this can be as easy as it gets. I also agree with you on the limitations and challenges that we face, but I also see much potential here (maybe beyond AdvancedGet if we continue to share)...

I suspect that for the XMAS threshold you use a slope calculation (i.e. <10, >20, etc.). Certainly would be interested to see the code and maybe collaborate on a few ideas. I would also like to get your input on expanding beyond the MAs and incorporating the other indicators as a means to counter some of the limitations you mentioned.

After playing with the code to simply change some of the "visuals", I can say that one can make a fun project out of this. From the look of things, Avery seemed to have a blast!



Russ