Repainting: what can you do agaisnt it???
> what can you do against repainting of indicators
Not much I'm afraid..
I sympathise with the frustration of repainting indicators - many times I think I have the 'holy grail' of signals only to get flipped in & out of positions, usually late AND wrong :(
To generalise, the 'better' an indicator looks, the more I suspect it repaints or at least lags - there are many culprits: Gann, Alligator, Heikin Ashi, HMA, STC, LaGuerre, Fractals, ZigZags, etc, etc
They have their place but more for fundamental position traders rather than swing or intra-day.
To clean up the code & reduce the latency, the very nature of the indicator would be changed and would not be 'the same thing'
For realism, you cant beat the 'moving average of the indicator'
Apply a Moving Average, typically Smoothed, dragged on to an existing indicator window
Set to the 'First Indicator' rather than the Open/High/Low/Close and see what happens
Works great on RSI, CCI, Momentum, OBV & DPO which can be a tad tricky to read on their own
Even better, apply two MA's and now you're getting somewhere
Try GBPJPY on the H4 or H1 with CCI (120), Smoothed MA (8) and Smoothed MA (4), both applied to 'First Indicator'
Looking mostly for supporting alignment in the early stages of a move and for divergence at the end
Clarity with concurrency ;)
FWIW
-BB-
What you can do? Throw it in the trash where it belongs, they're totally unsuited as tradesignal indicators!
I can't understand why people go to all that effort to create all these really nice looking repainting indicators,
when all they really needed to do is to apply a left shift to an MA....
So if you come across a promising indicator that looks like it would help you in making trade decissions, first
thing to do is check if it is repainting... it it does just discard it.
> what can you do against repainting of indicators
Not much I'm afraid..
I sympathise with the frustration of repainting indicators - many times I think I have the 'holy grail' of signals only to get flipped in & out of positions, usually late AND wrong :(
To generalise, the 'better' an indicator looks, the more I suspect it repaints or at least lags - there are many culprits: Gann, Alligator, Heikin Ashi, HMA, STC, LaGuerre, Fractals, ZigZags, etc, etc
They have their place but more for fundamental position traders rather than swing or intra-day.
To clean up the code & reduce the latency, the very nature of the indicator would be changed and would not be 'the same thing'
For realism, you cant beat the 'moving average of the indicator'
Apply a Moving Average, typically Smoothed, dragged on to an existing indicator window
Set to the 'First Indicator' rather than the Open/High/Low/Close and see what happens
Works great on RSI, CCI, Momentum, OBV & DPO which can be a tad tricky to read on their own
Even better, apply two MA's and now you're getting somewhere
Try GBPJPY on the H4 or H1 with CCI (120), Smoothed MA (8) and Smoothed MA (4), both applied to 'First Indicator'
Looking mostly for supporting alignment in the early stages of a move and for divergence at the end
Clarity with concurrency ;)
FWIW
-BB-
BarrowBoy
Thanks for the link to the youtube video!! I did not know that you could make and place an indicator on an indicator. I can totally see how this would work very well as an indicator!!!
Of course this leads to the follow on question: how do you code the indicator on indicator action into a usable EA? I get how to code the indicators "directly", but am challenged to see the coding of this. Do you have a link to another posting where you have provided direction to this on the forum in the past?
Thanks!!!
CountingMan
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Hi,
what can you do against repainting of indicators? When you do a visual mode backtest of an empty EA and you drop the indicator on the chart, you really see how it performs (repainting or not),
opposed to simply putting it on a live chart (too slow to evaluate repainting). So if you have seen that it does repaint, what can you do against that?
Is there a general rule that you need to follow? Or do you have to go through the code, step by step?
As an example I have attached a Hodrick-Prescott filter, an econometric method to split timeseries in a trend part and a cycle part. I like it a lot,
but only it repaints... How do I clean up this code so I get a "realistic" view of where it will have been on the past bars?
thx for the feedback guys, this forum is great!