Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
After watching Ehlers presentation one question came to mind:
1) Why would you extract the high frequencies to filter them afterwards?
2) Wouldn't it be easier to just apply the SuperSmoother directly to price and use it as input for Stochastics (or any other indicator) ?
I don't know if anyone has already tested and made a comparison between these two options. It is the underlying concept that I don't understand yet. Maybe someone with Signal Processing background has a good insight.
Amibroker 5.95 introduced the IIR (Impulse Response Filter), which should provide better performance than looping.
I am a rookie here so I can't provide a link to its readme -- too bad.
Sorry for 3y delayed response, I haven't notice that anyone response. So on your picture you compare SMA(l/2) and SS(l) because its the formula to calculate period parameter between them ?? (so eg SMA(5) and SS(10) )
The chart compares a SMA(20) to a 3-pole SuperSmoother(20).
However, the bar period has no meaning for many filters.
In order to find comparable bar periods you would need to calculate the lag of the SuperSmoother and compare it to the SMA.
This has been done for SMA and EMA.
Originally, Wilder's (exponential) average has been used with a smoothing constant k=1/N.
However, Wilder's average(N) is much slower than a SMA(N).
Therefore Jack Hutson suggested to use the formula k = 2/(N+1) instead of k = 1/N. (see article in one of the first issues of Stocks & Commodities).
With the introduction of the smoothing constant k = 2/(N+1) an EMA(N) had now a comparable lag to a SMA(N). The EMA is calibrated against the SMA.
This has not been done for the SuperSmoother. But the picture shows that the SS(20) has both a smaller lag and superior smoothing when compared to the SMA(20).
I think it just the best to evaluate its performance in trading systems as according to his presentation it should
clean data from aliasing noise and spectral dilation so after denoising SNR ratio should be higher and data much
predictable. Unfortunately its opposite.... more info here