Digital ACSTrend - page 11
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Beast of market never want traders to feel her pulse.
When I change to 20 Bars adpation of ftlm-stlm, a lot of red verticale line appeared on chart. The peak and trough of ftlm line can seen clearly follow price, but, it lags 3 to 5 bars to peak and trough on chart.
The length of cycle is often evolving with Fibonacci serie, I found this on most time frame of charts. So, we can validate and project the variation of cycle length in order to catch up the pulse of market beast.
MESA is very good, powerful tool. It can detect "PAST" cycle, frequencies, etc. But, it wouldn't detect and project the evolution of cycle. I may be wrong due to I am not expert in this field. I just watch chart, and get my feel.
When different period is used in Hurst indicator, the indicator may show difference reading. How can Hurst indicator be adapted? I put 3 periods on charts (attached).
Thank John,
I found your Mama2 and ftlm-stlm on higher time frame(D1,H4, H1) are very good, higher performance, wonderful work. Thx.
iVAR & FGDI
The iVAR calculates the variation index on a previous interval which is 2^n long. The "n" parameter is specified by the user.
So if you adapt the n parameter to the dominant cycle you may get nothing.
However the FGDI can be used with this methodology.
However I think that the fractal dynamics should be adapted and estimated according to the volatility. And for that we may need a tick chart. The bars are hiding a lot of useful information from time to time. You may get the signal too late. So iVAR as it never recalculate the current bar you will have your estimation at the close. The FGDI recalculates as any other indicator the last bar, so you have current information of what is going on. That is why you have the impression that the FGDI is more reactive.
By the way nice, nice color coding.
What ASCTrend you use on the shots? The ASCTrend should be adapted to the local conditions if you see three bad signals in a row that mens that something is wrong. It is just a tool after all. My recommendations for the standard ASCTrend are 8 to 10, sometimes go even to 16 to see the bigger picture.
Hi, John,
On the screen shot, I use ASCTrend_adaptive_cycleperiod. It is quite good. Some bad signal can be filtered by ftlm_stlm. I believe, most of trend signal need a filter to filte some fake or bad signal, then, get a good performance. I am still watching it.
there is a broad indicator saying ASC, show us the diagram and tell us whether it is accurate
Here I add an example what I mean.
We have several adaptive ASCtrend mods.
The classic adaptive digital WPR ASCtrend mod
When I use with adaptation to one full dominant cycle period I have very nice smooth results. A lot of noise is filtered. Use Nyquist of 1.0.
When I use with adaptation to a half of a period more signals are generated
Use Nyquist of 0.5.
The Adaptive JStochastic ASCTrend mod
The JStochastic mod is more reactive that the previous.
I prefer it with half a period.
When I compare the two mods. My observations are that the JStochastic is better with 0.5 Nyquist, and the WPR is better with 1.0 Nyquist.
MESA adaptive ASCTrend with WPR
This is something really reactive it tries to pick tops and bottoms but is easily misguided by the noise. However at high time frames it can give astonishingly accurate signals.
This is what I get with the default setup. Really the adaptive mods give you signals or they do not, I do not try to optimize.
On the shot I give a picture how it looks the analysis at the classic 1 h time frame. For day trading we need to analyse in full 15 m and 30 m time frame.
The Alma with caterpillar is an additional help as a filter. The Slope is the Alma the dots use the caterpillar so we can expect them to recalculate (repaint) from time to time (however they are usefull as a guide where is the momentum).
Please feel free to share your views and shots.interesting
Hey John
I begin to understand your philosophy
For me, the IVAR and FGDI are very important. Off course the asctrend mods are also unvaluable !
When the FGDI are in range mode and horizontal, i like to trade the extremums of the adaptive channel you post before.
Thanks for all
Some examples
Here I add an example what I mean.
We have several adaptive ASCtrend mods.
The classic adaptive digital WPR ASCtrend mod
When I use with adaptation to one full dominant cycle period I have very nice smooth results. A lot of noise is filtered. Use Nyquist of 1.0.
When I use with adaptation to a half of a period more signals are generated
Use Nyquist of 0.5.
The Adaptive JStochastic ASCTrend mod
The JStochastic mod is more reactive that the previous.
I prefer it with half a period.
When I compare the two mods. My observations are that the JStochastic is better with 0.5 Nyquist, and the WPR is better with 1.0 Nyquist.
MESA adaptive ASCTrend with WPR
This is something really reactive it tries to pick tops and bottoms but is easily misguided by the noise. However at high time frames it can give astonishingly accurate signals.
This is what I get with the default setup. Really the adaptive mods give you signals or they do not, I do not try to optimize.
DO NOT FORGET YOU CAN VARY min_period parameter, the default is set to 5 that is why it is so reactive, if you set to 20 the results will be much smoother.
On the shot I give a picture how it looks the analysis at the classic 1 h time frame. For day trading we need to analyse in full 15 m and 30 m time frame.
The Alma with caterpillar is an additional help as a filter. The Slope is the Alma the dots use the caterpillar so we can expect them to recalculate (repaint) from time to time (however they are usefull as a guide where is the momentum).
Please feel free to share your views and shots.
MESA Mod
Thank you for the comments.
I just want to mention something important.
DO NOT FORGET YOU CAN VARY the min_period parameter, the default is set to 5 that is why it is so reactive, if you set to 20 the results will be much smoother.
I have prepared some shots just to give you the idea.
I fact this is the lowest limit of the mesa cutoff_frequencies. If you are experienced you can make manual adjustment of the necessary minimum cut-off frequency using the MESA spectral analyzer. If someone is experienced with that please do not hesitate.