TrendNoTrend Indicator

TrendNoTrend Indicator

The “Linear regression slope” indicates the direction and intensity of the current trend.
In the abstract hypothesis of total absence of trend the slope of the linear regression is equal to zero: the linear regression is a line parallel to the abscissa axis.
A statistical way of establishing that there is no trend, and so we are in a laterality phase of the market, consists in establishing that the slope of our linear regression is statistically equal to zero in the observation period.
To do this in statistics we use the “t-test”.
For all those who love statistical problems I attach the pdf files with the related explanations.
The statistical formula is translated into the following code.
In a normal distribution, 95% of the data is included in a confidence interval of “Average + – 1.96”.
A “t-distribution” with infinitely many degrees of freedom is a normal distribution.
With an approximation we can state that if    t <= 1.96  then our LinearRegressionSlope is equal to zero = there is no trend.

 

Share this

Risk disclosure:

No information on this site is investment advice or a solicitation to buy or sell any financial instrument. Past performance is not indicative of future results. Trading may expose you to risk of loss greater than your deposits and is only suitable for experienced investors who have sufficient financial means to bear such risk.

ProRealTime ITF files and other attachments : How to import ITF files into ProRealTime platform?

New! PRC is also now on YouTube, subscribe to our channel for exclusive content and tutorials

  1. Pepsmile • 309 days ago #

    Hi FULVI,
    Interesting. But seems to take into account only bull trend.

  2. Jan Wind • 304 days ago #

    Dear Fulvio09,

    The concept of calculating trend or no trend attracts me much.

    First question: the linear regression slope indicator itself gives already the strength of the trend in itself, what does your calculation add to this ?

    If I compare the graph of the standard linear regression slope indicator with your indicator, and plotting your defined t-values as indicator, (instead of making them 0 or 1), the t-values seems just a multiplication of the linear regression slope values.
    Therefore my second question: could you explain your formula shortly per item, why it is used , and with a consistent number example ? (For example one of the items STD[AA](Barindex) why is this used ?)
    Maybe the standard indicator R2 can be used ?

    Thanks in advance, Jan

    • FULVIO09 • 299 days ago #

      I try to answer even if, not being a statistic, it is not easy for me to do it, and even more in these few lines.
      The “t-test” is used in statistics to compare two groups of data.
      Basically it compares the difference between the mean of two groups, considering the desperation of the data (expressed as standard deviation).
      The result “t” of the “t-test” is compared to a table of the values of the “t-test” (see: ” https://www.itl.nist.gov/div898/handbook/eda/section3/eda3672.htm ” ).
      In the table for each degree of freedom (first column of the table) you will find next to, the values corresponding to the probability you want to check.
      If you compare, for example, two groups with 10 data each, the degrees of freedom are 10 + 10 – 2 = 18.
      If you compare a large number of data (degrees of freedom> 100 to infinity) (last row of the table) you find that for a value of t <1.96 there is a 95% probability that the two groups of data are equal.

      The "t-test" formula applied to the specific case of "slope of a regression line" can be found at the end of the attached "t-test" pdf file.
      In essence it compares the difference between the slope of two linear regressions considering the dispersion of data (expressed as "standard error of the slope coefficient").
      I suggest you to watch: " https://www.youtube.com/watch?v=j5oPzAJvnVI " you will need to understand why the standard error is also influenced by "STD [AA] (Barindex)".

  3. cfta • 292 days ago #

    Hi Fulvio, many thanks for sharing this clever indicator, much appreciated. As Pepsmile mentioned it seems that it only accounts for bull trend and ignores bear trend, I modified it in order to include the bear trend. 1 = bull trend, 0 = no trend and -1= bear trend. Enjoy!

  4. cfta • 292 days ago #

    Sorry adding the PRT to code didn’t work in the above comment so I guess we will have to settle for regular text format…

    // period da ottimizzare
    AA = period
    t=(LinearRegressionSlope[AA](close)-0)*SQRT(AA-2)/(STE[AA](close)/STD[AA](Barindex))
    if t>1.96 then
    beta=1
    endif

    if t<-1.96 then
    beta=-1
    endif

    if t-1.96 then
    beta=0
    endif

    return beta

  5. juanj • 217 days ago #

    Something appears to be wrong with this formula as graphing t shows t to almost never be within the -2 to + 2 range. The only reason I believe the author sees extended periods of beta = 0 is due to the fact that he is including bearish linear regression < -1.96 as beta = 0

avatar
Register or

Likes

avatar avatar avatar avatar avatar avatar avatar avatar
Related users ' posts
swapping Excellent Nicolas ;)
Violet Hi Nicolas, There seems to be a small error in the code, which causes the channel not to...
Nicolas Hmm, I don't think so, otherwise the channel will have a moving start and not anchored anymo...
Juanjo Muchas gracias por la explicación, muy interesante.
vandenacker THANKS A A LOT FOR SHARING. I like your trade. I try to remproduce your board but i dont arr...
umebon Tnx for sharing!! Can someone help me how i can call for the mydayhigh and mydaylow for to u...
coscar Ottimo lavoro. come sempre!
luxrun Nello studio di Sepiashvili viene descritto anche un altro indicatore, il Q-indicator, che è...
Nicolas Just add a linear regression moving average of 200 periods on the chart and you'll get the s...
Partha Banerjee Hello Nicolas, this is great, however the actual PRT standard error and standard deviation c...
Nicolas If you want to anchor the channel to a specific candlestick, that would be possible but with...
sharksteven Hi BALMORA, looks like a good compliment algorithm. I tried to test it on the EUR USD but it...
kompiuter Hello, how did you optimized all these variable? If I try to optimize only two variables it...
rrenaud_ig Great work ! My only concern is about risk management. Your results are based on 10k. Max lo...
Paul in ProOrder support, at the bottom
Maik2404 wie kann ich den Code bekommen Paul?
jens_kittner Works with US Crude at 1h as well!
osupero Buenisimo Sr Nicolas, seria posible que sólo marcara por encima o por debajo de una media mo...
osupero https://www.screencast.com/t/1FD7s0YdZ......DE NUEVO MIL GRACIAS
capgros Bonjour @Nicolas, Thanks a lot for this tool, it is very useful for me. I would like to s...
Nicolas
9 months ago
BSTrend
v10.3
BSTrend
8
Indicators
JvdG I meant of course: since ld56 is one of the constituents of ld48;
Nicolas You are right about ld56. Since this is a translation from another programming language, I d...
Florian Legeard C’est à dire sur le prix et non sur le temps ? Merci
Nicolas
11 months ago
Maxime Baudin Interesting, Thank you! :)
franck-david Bonjour Nicolas , actuellement je me forme grace a vos videos tres tres bien expliques mer...
Nicolas Merci. ça n'est pas le lieu pour poser ce genre de question :) Merci d'utiliser le forum Pro...
Nicolas Agréguelo a la tabla de precios usando el botón de configuración de precios en la tabla.
Traderfox06 Dear Nicolas, I believe in the high potential of quadratic regression for an automated sy...
Nicolas S11 variable is not a loop but a ratio calculation. In real time the codes is processed on e...
Bard Thanks Gabri, How do you like your Turkey cooked? https://www.dropbox.com/s/1afg6qqq...
gabri Brined :) look on this indicator how on the 6th and on the 9th you have very high positive s...
Pepsmile Hi everybody, An application could be found by applying this indicator on the strategy prof...
Canale di Fill By Filippo si si... farò un video a breve
AngelPujalt ok. Gracias.
Stenozar Hi Filippo, any news about your video to explain how to use this indicator? thanks
Vonasi Discussion on this indicator/filter can be found here: https://www.prorealcode.com/topic/121...
pableitor Hola Domenec, un indicador muy original y util que ciertamente añade mucha informacion a la ...
Domenec Hola Pableitor porque aqui lo que pretendo es programar un nuevo codigao mas ajustado para q...
Juanjo Hola Domenec. He visto en los videos que hablas continuamente de mercado de futuros con grá...
Marcot18 Buonasera Maurinho, interessante l'indicatore che ha i proposto ma ancora di più interessant...
maurinho ciao marcot18 l'indicatore pubblico /proffessional lo puoi scaricare si chiama progo di will...
james effiong This strategy looks good
Geronima Ortiz I watched the video three times, but I can not put the npips parameter as a variable. I do...
Geronima Ortiz I think the video is for an old version of prorealtime, the images do not correspond at all ...
Nicolas indicator1 = CALL “PRC_StopReversal”[npips, 1] and define npips in the optimization window ...

Top