bluecrow 

Thanks Fat Tails, appreciated it! Thanks for including the screenshot. I see there's a trade-off between setting the smoothing filter higher which increases the lag and leaving the default parameters as is. Are there any practical solutions that would reduce the lag inherent in the supersmooth filter? Many thanks, Joe

If you increase the period of a moving average, it becomes smoother and you increase the lag. If you decrease the period both smoothness and lag are reduced. For a SMA or an EMA the typical lag is about half the period.

There is always a trade-off between smoothness and lag-reduction. You cannot have both. If you set the period to 1, the lag is completely gone, but there is no smoothing.

Fat Tails 

If you increase the period of a moving average, it becomes smoother and you increase the lag. If you decrease the period both smoothness and lag are reduced. For a SMA or an EMA the typical lag is about half the period. There is always a trade-off between smoothness and lag-reduction. You cannot have both. If you set the period to 1, the lag is completely gone, but there is no smoothing.

Ehler SuperSmoother works very similar like ordinary SMA so there is no superiority here. More info with pictures here

All John Ehlers Indicators... - Page 59

Krzysztof

krzysiaczek99 

Ehler SuperSmoother works very similar like ordinary SMA so there is no superiority here. More info with pictures here All John Ehlers Indicators... - Page 59 Krzysztof

@krzysiaczek99: If you tell me that a pair of scissors is superior to a pencil, I would certainly respond that my preference depends on the case.

The SMA is a finite response filter which is not really smooth. As all finite response filters it is slow to digest rapid moves or gaps.

The Supersmoother is an infinite response filter, which is generally faster to react and considerably smoother compared to the SMA.

I do not agree that the SuperSmoother has a similar behavior as the SMA, see chart below:

PurelyRandom 

Hi, I am having trouble with some super smoother code for Amibroker - it seems to adjust the amplitude of the orgininal arrray to be smoothed, which I don't believe should be the case? Here is my indicator code...unless I have made a msitake somewhere along the line, here is the afl: function SuperSmooth(arr,period) { Filt = a1 = b1 = c1 = c2 = c3 =0; a1 = exp(-1.414*3.14159 / period); b1 = 2*a1*cos(1.1414*3.14159/period); c2=b1; c3 = -a1*a1; c1 = 1 -c2 - c3; Filt = (c1*(arr + Ref(arr,-1))/2) + c2*Ref(Filt,-1) + c3*Ref(Filt,-2); return Filt; } smoothedClose = SuperSmooth(C,10); Plot( smoothedClose, "smoothedClose", colorRed, styleLine );

Maybe you already figured out that the Filt line is wrong. Since it depends on previous values of Filt, you should use a loop.
i.e.
for ( i = 2; i < BarCount; i++ )
Filt[i] = ( c1 * ( arr[i] + arr[i - 1] ) / 2 ) + c2 * Filt[i - 1] + c3 * Filt[i - 2];

There is also the article in January 2014 Technical Analysis of STOCKS & COMMODITIES
"Predictive And Successful Indicators" by John F. Ehlers, Ph.D.

Code for various platforms are in the Traders Tip section.
Specifically the Amibroker code for SuperSmoother function can be found here:
TRADERS TIPS - January 2014

John

John Ehlers is back on Thursday, June 5th:

Webinar: John Ehlers on Anticipating Turning Points Hi guys, It is my pleasure to welcome back John Ehlers for another webinar on Thursday, June 5th at 4:30 PM Eastern US. The bullet points include: • Market Data is Fractal – Longer waves have bigger swings than shorter waves. • Spectral …

Be sure to attend!

Mike

Fat Tails 

@Big Mike: Unfortunately I did not have the time to attend the webinar this evening, although I admire the work of John Ehlers. I am the proud owner of one of his books, and have read many articles he has published in Technical Analysis of Stocks & Commodities. I also recommend to download the technical papers, which can be found on his website: Home There are numerous indicators available for NinjaTrader that are based on the work of John Ehlers. This is just a brief summary. .......... All of this adds up to an impressive collection of tools for technical analysis. John Ehlers, thank you very much for your contributions!

Hi FT,

First, thanks for such a nice compilation if indicators based on John's work.
In the download section you specify that some or most of these were originally uploaded by Piersh.
Were these easylanguage by any chance?
I would like to know if there are versions of these for TradeStation/Multicharts.

Thx

andby 

Hi FT, First, thanks for such a nice compilation if indicators based on John's work. In the download section you specify that some or most of these were originally uploaded by Piersh. Were these easylanguage by any chance? I would like to know if there are versions of these for TradeStation/Multicharts. Thx

John Ehlers has originally coded is indicators for TradeStation. It should therefore easy to find them somewhere. The collection by @piersh was coded for NT 6.5. I have just made a few modifications to allow to use them with NT7. So you may not consider me being the author of that package.

The TradeStation code for many indicators is also printend in John Ehlers' books, and some of them are available on his website.

Home

Found this link which is pretty interesting for MC users:

MultiCharts: Trading Software for Automated Trading and Backtesting ? View topic - Collections of EasyLanguage Scripts

Following indicators are contained in the attached:

JohnEhler 2-pole Butterworth filter
JohnEhler 3-pole Butterworth filter
JohnEhler 3-pole Super Smoother filter
JohnEhler Adaptive CG
JohnEhler Fisher Cyber Cycle
JohnEhler Instantaneous Trend
JohnEhler Leading indicator
JohnEhler Sine Wave
JohnEhler 2-pole Super Smoother filter
JohnEhler Adaptive RVI
JohnEhler Adaptive Cyber Cycle
JohnEhler Center of Gravity
JohnEhler Cyber Cycle
JohnEhler Cycle Measuring
JohnEhler Cycle Period Measurement
JohnEhler Fisher Stochastic CG Oscillator
JohnEhler Fisher Transform indicator
JohnEhler Laguerre Filter
JohnEhler Laguerre RSI
JohnEhler MAMA FAMA
JohnEhler Relative Vigor Index
JohnEhler Sine Wave
JohnEhler Sine Wave2
JohnEhler Smoothed Adaptive Momentum
JohnEhler Stochastic CG Oscillator
JohnEhler Stochastic Cyber Cycle
JohnEhler Stochastic RSI
JohnEhler Stochastic RVI
JohnEhler Smoothed Adaptive Momentum Strtategy
JohnEhler STOCHASTIC RSI STRATEGY

I came across Ehlers' SuperSmoother and Roofing filter code for various platforms (Tradestation, Ninjatrader, eSignal, etc) at TRADERS’ TIPS - January 2014

there was also mistake in b1 calculation, sqrt(2) was typed 1.1414.

Here is working indicator for Amibroker after all corrections

function SuperSmooth(arr,period)
{

Filt = a1 = b1 = c1 = c2 = c3 =0;
a1 = exp(-1.414*3.14159 / period);
b1 = 2*a1*cos( (1.414*3.14159/period) );
c2=b1;
c3 = -a1*a1;
c1 = 1 -c2 - c3;

for ( i = 2; i < BarCount; i++ )
Filt[i] = ( c1 * ( arr[i] + arr[i - 1] ) / 2 ) + c2 * Filt[i - 1] + c3 * Filt[i - 2];
return Filt;
}
smoothedClose = SuperSmooth(C,10);
Plot( smoothedClose, "smoothedClose", colorRed, styleLine );