If you Look at my chart ,upper left corner,you will see GBPJPY M2...;)..Why is that so?....Because I use Constant Range Bars..why?..Because as you have had the opportunity to experience ,they smoothe the signal and decrease the noise....Without delay....without repainting.
I have found that 99% of indicators do not work,and the other 1% that DO work,they work better in CRBs,if you know how to size them....So,Igorad and mladen indicators deserve to be put on CRBs
Now,the 2 questions you may have....;)
1-What do I use to create the CRBs?...I use a commercial product(a script) by mql services,whose owner is a member of this Forum(Michal or Michel)...since they are one of the few commercials that deliver,sorry,i won`t post the script here,the script is affordable,and there are,probably,several similar scripts or indicators,for free.
2-How do I size them?...I use ATR(100)..and I choose the ,rounded,number of pips for H4 ATR(100),which for GBPJPY is 100 pips,so the bars in that chart are 100 pips from high to low...For EURUSD I use 50 pips and for other pairs I size accordingly....Why Do I use h4?...Because my methods seem to work better in h4(and EA tests usually confirm that),so,I thought that equalizing the bars for h4 volatility will improve them even more...It did.
In any case,I combine techniques with CRBs and techniques with several normal timeframes,so,don`t worry,I too use normal charts,and will use in my promised thread,but CRBs afford me a better "less biased" view...
Saw your M2 chart was guessing that's what it was but have Renko script, very ingenious use of these, kind of quickly threw this together using TMA Bands with Renko, looks like a lot of opportunities.Looking forward to your thread.
ps) bands aren't set right just threw this together real quick before going to work!!
Just as a temporary help : in this one those ATR multipliers and ATR length can be set from parameters
One more thing (something I usually do not do, but) : Michal's CRB is a very professionally made script. Also mrtools's solution is elegant as usual
I haven't notice you were using Renko untill now. I found it an interesting mather to explore since i also think 99% of the indicators doesn't work, they all lag and are all affected by the time factor. By eliminating the time i found that charts are more precise and indicator more effective, but, now the Renko box is becoming my main concern, your method of defining bar size is interesting. but box size defines a chart so much and is such an important factor that i still hesitate in using CRBs. I am looking for more info on that mather on how to define box size. I think that optimizing box size is a University thesis by itself. All you inputs will be deeply consider. I personally use a free script. Here it is.
Mladen:Thanks for the new bands,they are more than fine...as usual
GreatYves:Thanks for the Renko script,it doesn`t do exactly the same as Michal`s one but it will more than suffice ...
You know,you are right when you say that the range of a CRB could be the subject of a university thesis...basically,one day I started thinking about...what is the right timeframe?Well,my systems usually work better if I start with H4 as a framing timeframe within which I can then mtf,but I know people that can and do successfully scalp the ES using just m5,m1 and then a tick chart...so,it is obvious that the right TF is not necessarily H4.
Additionally if you read the theories of Trader X(Jason Alan Jankowsky previous nom de plume)about time compression,that basically say that,in regards to trading, 5 minutes are not just 5 minutes...because it is obvious that 5 minutes at the Asian session are not the same as 5 minutes at NFP announcement...so,if we want to "normalize" time in order to compare different periods,we must do so using volatility as the equalizing factor...and then we will be able to understand that 5 minutes at NFP are equal to 4/8 hours at the Asian session....This led me to think...So,we can`t attach the same importance to both M5 periods...BUT...that is exactly what we do if we use a M5 chart,we say that 5 minutes of NFP time are equally relevant as 5 minutes at the Asian session,and this is conceptually wrong,so,our way of charting should be changed.
These theories were very interesting but still didn`t answer my query for the right timeframe,even if they provided another piece of the puzzle(volatility).
Then ,one day, I realized that the right timeframe was no timeframe...If your system works well with H4 use CRBs equalized by the ATR of H4,if your system works well in m15,do the same with CRBs equalized by the ATR of M15...and your results will improve...Why?Because what you have effectively done is to denoise(partially) the timeseries ,with 0 lag,creating a better scenario in terms of Signal to Noise for whatever was working before,to work better now.
The actual stage of my thinking is to test the use Multi CRBs..4 screens each with CRBs equalized to the ATRs of H4,H1,M15,M5 and see if this adds some useful information .
BTW..H4 ATR (100)of GBPJPY has gone from 102 to 79,and we have another 2B buy signal(penetration of previous bottom plus up close plus a close with breakout of the high of the previous bar)in the making(still not completed),so,if the signal happens,it is very probable that we have seen the bottom at 148.58...when volatility compresses,time expands ,and the direction of the previous move is prone to change.
Next week I will start the thread.
Simba, your realization is as enlightening as it is logical. Brilliant.
Looking forward to your thread.
Have a great weekend all. No trading today for me when four witches are flying by...
I totally agree, Simba's realization is brilliant indeed. Thank you so much for the information about CRB.
Hopefully this is the right place to post this, if not could someone please point me in the right direction.
Could someone please mod the indicator Bollinger Band %b (the one I am using is made by Linuxser) for me.
1) Make the price and mode used for the bollinger band calculations chooseable.
2) Have a input to smooth the curve using a moving average (where 1 would be no smoothing)
3) Have it so several (up to three) different periods can be shown in the same chart window, the scale on the window would autosize to the size needed for the largest value.
4) Alerts when a userselectable value is crossed by one or more of the curves (i.e. crossing down over 0.8)
once again thanks for your insights. When you're testing, you use Michal's CRB script you've been talking about?
Nothing to do with the previous ones yet (just something I stumbled upon by the way and it seemed interesting)
A description from a Markos Katsanos book (Intermarket Trading Strategies) :
• It is self-contained in the sense that it eliminates the need for a second indicator to identify the trend direction.
• It is better in identifying trend reversals.
• It provides more accurate readings in cases of temporary pullbacks.
PS: signal line is an addition (does not exist in the original) To turn it off just use SignalPeriod< 2.
Thanks for your kind words,yes,I will test using Michal`s script,though the same tests could be done with Renko script,and probably similar results.