Sigma Bands for CQG

[Soultrader 12]

Not an indicator based trader but received this a while back and may be of interest to some. The indicator plots Sigma bands. (please dont ask what they are) Formula is not closed so after you import you can take a look. Snapshot is attached.

 

Sigma.pac

[metalhead 10]

why can't we ask what they are?

[brownsfan019 17]

why can't we ask what they are?

 

Yeah, why can't we ask?

 

:rofl:

 

I believe he said don't ask.

 

So you ask.

 

:crap:

[Soultrader 12]

I have no idea what they are... I dont use them. They were given to me by an institutional trader.

[waveslider 10]

Sigma bands are related to a standard deviation from a mean, usually a moving average (why not). The example has a very tight average and normally the average is wide, so that if price reaches 2 std. devs. then supposedly you have a 90% chance of a bounce. Why? It's stretched too far.

 

How many have tried to find indicators that show price has stretched too far? People trying to catch tops and bottoms. They go out of business quick.

 

Hurst was the one that made these famous, and I haven't read Hurst so feel free to correct me.

[metalhead 10]

So they're the same as Bollinger Bands then?

[waveslider 10]

OK correction, I got my indicator out and looked at it. Its a standard error around a linear regression average. So definitely different. If nothing else they look pretty.

[davla 10]

is it possible to post me a .txt file of this, as I dont use CQG and would really like to review this method.

 

Thanks very much.

 

If there is anything I can do for anyone please let me know

rgds

Davla

[davem1979 11]

Recently I have been looking into this and I found some links online that explain more about what the Sigma Boundary is:

 

https://secure.barchart.com/crb/images/emini/CRB-EMini-sample.pdf

 

http://bcs.barchart.com/pdf/MFG_OptionsResearch2.pdf

 

http://www.oir.com/sample/FX-VL-CH.PDF

-----------

I started thinking about using different calculations to find "statistical extremes" from Brett Steenbarger. He says that he uses volatility adjusted pivots. Maybe he uses ATR, or maybe even something similar to the Sigma Boundary using Implied Volatility in his pivot calculations. http://traderfeed.blogspot.com/2009/08/volatility-adjusted-price-targets-for.html

----------

For the Sigma Bands, my initial thought was to plot these bands around either the typical price, the close, the open of the new day, or even the prior days VWAP http://www.precisioncapmgt.com/2009/09/30/using-prior-days-closing-vwap-as-supportresistance/

 

I think the original intention of the Sigma bands is to just look at closing prices, so they say that 68% of the time the next days price will close within the 1st Standard deviation, and 95% of the time, the next days price will close within the 2nd Standard deviation etc.

 

I think it will be very interesting to look at the Sigma Boundary vs. using an ATR to define "statistical extremes". I like the idea of the Sigma Boundary because it seems to me like it's a "market internal" since options traders are always in tune with Implied Volatility, since it is used for the pricing of options.

 

I could see using the Sigma Boundaries similar to the 10 day ATR, when given the market structure, internals, volume, one can look to fade them or look for breakouts.

 

Thoughts?

 

All the Best,

David

[davem1979 11]

Here's another article I found that goes further in depth about how to calculate the daily range implied by volatility.

 

http://www.ivolatility.com/news/Volatility_to_work.pdf

 

Check out the formula on page 7 and page 8.

 

Also, as most people seem to use 252 days, some also use 365 days, so it might be worth experimenting a little a see what makes sense to use. The same goes for plotting the standard deviation around the closing price or the opening price and seeing which seems more accurate.

 

Best,

David

[davem1979 11]

I watched a video of Dean Mouscher's I found online and he mentioned he spoke with Robert Whaley, the guy who developed the VIX for the CME in 1990, and he got an equation from him for calculating the Median Daily Range for the S&P 500 as implied by the VIX .

 

That equation is: (% Implied Volatilty/SQRT 365)*(Stock Price)*2*.675

 

I am pretty sure the 2*.675 is the IQR which is explained here: Interquartile range - Wikipedia, the free encyclopedia

 

Best,

David

[shrike 10]

Hi to everyone

 

I 've found this formula posted by TRO:

 

 

[LegacyColorValue = true]; 

{_SIGMA3  }


{Compliments of Avery T. Horton, Jr. aka TheRumpledOne 

I hope you find this program useful and profitable.

DONATIONS AND GIFTS ACCEPTED 

P O BOX 43575, TUCSON, AZ 85733 }

{ © Copyright 2006 Avery T. Horton, Jr.}



Inputs: 

iCalcType( "D" ), 		// D = Dynamic, S = Static - at start of day, T = at specified time
iCalcDate( currentdate ), 
iCalcTime ( 0000 ), 
iPrice(  close ),	
iDecimals( 2 ),

iLength( 20 ),
iDev1( 1 ),
iDev2( 1.5 ), 
iDev3( 2 ),


iDisplace( 0 ),

iAvgColor( white ),
iDev1Color( yellow ),
iDev2Color( green ),
iDev3Color( blue );

variables: 

Avg( 0 ), 
SDev( 0 ), 
xSD3( 0 ), 
xSD2( 0 ), 
xSD1( 0 ), 
msg(""), 
ret_str(""), 
ret_val(0);


variables: 

sCalcSwitch(false),
sPlotSwitch(false),

xSigma1( 0 ),

xSigma2( 0 ),


xLengthInMinutes(0),	


xPeriods(60),
xInterval(0),

sFirstPass(true),

xMult( 0 ),



FG1(white),
BG1(black),

fg2(white),
bg2(black),

fg3(white),
bg3(black),

fg4(white),
bg4(black),

vdummy("");


{commentary variables}

variables: 
xcomm(0),
oComm1( "" ), 	
oComm2( "" ), 	
oComm3( "" ), 	
oComm4( "" ), 	
oComm5( "" ),
oComm6( "" ),
oComm7( "" ),
oComm8( "" ),
oComm9( "" ), 	
oComm10( "" ); 

{first time through}

if sFirstPass
then begin

sFirstPass = false;

{bar test}

If bartype = 4
then xInterval = 94
else
If bartype = 3
then xInterval = 93
else
If bartype = 2
then xInterval = 92
else
If bartype = 1 or bartype=5
then begin
xInterval = BarInterval;
end; { If bartype = 1  }

end; {if sFirstPass}

{ BEGIN PROCESSING }

sCalcSwitch = false;

If ( ( iCalcType = "S" or iCalcType = "s" ) and d <> d[1] )
or ( iCalcType = "D" or iCalcType = "d" )
or ( ( iCalcType = "T" or iCalcType = "t" ) and time this bar = iCalcTime )
then sCalcSwitch = true;

If sCalcSwitch
then begin

{ INITIALIZE }


FG1 = WHITE;
BG1 = BLACK;

fg2 = WHITE;
bg2 = BLACK;

fg3 = WHITE;
bg3 = BLACK;

fg4 = WHITE;
bg4 = BLACK;



{ MINUTE PROCESSING }

Value90 = TimeToMinutes(0001);
Value91 = TimeToMinutes(2359);
Value92 = Value91 - Value90;

Value80 = MinMove/PriceScale;

xLengthInMinutes = IntPortion(Value92 / BarInterval);


{ calculations }

Avg 	= AverageFC( iPrice, iLength ) ;
SDev 	= StandardDev( iPrice, xLengthInMinutes , 1 ) ;

if iPrice > Avg
then xMult = 1
else xMult = -1 ;

xSD1 = Avg + iDev1 * SDev * xMult ;
xSD2 = Avg + iDev2 * SDev * xMult ;
xSD3 = Avg + iDev3 * SDev * xMult ;

end;  // If sCalcSwitch


{ plots  }

If( ( iCalcType = "T" or iCalcType = "t" ) and time this bar < iCalcTime )
then sPlotSwitch = false;

If ( ( iCalcType = "S" or iCalcType = "s" ) and d <> d[1] )
or ( iCalcType = "D" or iCalcType = "d" )
or ( ( iCalcType = "T" or iCalcType = "t" ) and time this bar = iCalcTime )
then sPlotSwitch = true;


If sPlotSwitch 
and d = iCalcDate 
then begin

Plot1( xSD1 , "SD1", iDev1Color ) ;
SetPlotBGColor( 1, bg1);

Plot2( xSD2 , "SD2", iDev2Color ) ;
SetPlotBGColor( 2, bg2);

Plot3( xSD3 , "SD3", iDev3Color ) ;
SetPlotBGColor( 3, bg3);


Plot4( Avg , "Avg", iAvgColor ) ;
SetPlotBGColor( 3, bg3);


end; // If sPlotSwitch 



{Found this:

Bollinger Bands

Description 
The Bollinger Bands were introduced by J. Bollinger. 

They provide a visual channel of upper and lower bounds that prices tend to stay between. 

The channel calculation is based on variation about a statistical mean over a certain look back period.

The channels are defined by the calculation of the standard deviation [b][red](sigma)[/red][/b] of the input value. 

The upper band is some multiplication factor of Sigma added to a simple moving average of the input value 
for the same period as the Sigma calculation. The lower band is the value minus Sigma times a multiplication factor.

The SMA is plotted as a dashed line and the upper and lower Bollinger Bands are plotted as thin, solid lines. 

A text note is added to the plot in the upper left corner showing the look back period and the upper and 
lower Bollinger Band Values. 

in here:

http://www.stockstoshop.com/bollinger.htm

}



// https://www.tradestation.com/Discussions/Topic.aspx?Topic_ID=31077



// value1  = StdDev( iPrice , iLength )  ;


//	value1  = StdDev( iPrice , iLength ) * ( iPrice * Volatility( iLength ) ) ;


//	value1  = StdDev( iPrice , iLength ) * Volatility( iLength ) ;

//	value1  = StdDev( iPrice , iLength ) + Volatility( iLength ) ;

//	xSigma1 = value1 * iLevel1 + iPrice ;

//	xSigma2 = value1 * iLevel2 + iPrice ;

// value2 = iPrice + iPrice * Volatility( iLength )  ;

// value1  = StdDev( value2 , iLength )  ;

[davem1979 11]

Hi Guys,

 

I think a discussion on the Sigma Boundary would be better served by moving to the Market Statistics forum. I am going to start a new thread over there that starts with a discussion of the Sigma Boundary and the probability cone as published in this PDF:

 

https://secure.barchart.com/crb/images/emini/CRB-EMini-sample.pdf

 

I think any other discussions of the Sigma Bands for CQG should continue in this thread.

 

Best,

David

[johnnydaymon 11]

The last Part of this code in the text file in post # 13 does not belong in the _sigma3 indicator! TRO's code Works on OHLC Bars,HLC Bars only. Doesn't work on Renko or daily charts, just minuts and tick charts.When Vwap is overlayed on the chart you see the white line and red line match. The white line is from sigma3 and the red is vwap. Now you know what the white line is.

_SIGMA3_REVISED.txt

_SIGMA3.ELD

Edited February 27, 2013 by johnnydaymon

[johnnydaymon 11]

Check out three different types all doing the job!