[VSA] Volume Spread Analysis

[VictorH]

Hello rayl54292,

This is my first post on this forum.
I was interested in the chart in your last post [3] and your comment about “input and brutal honesty�

This is my analysis of VSA in terms of looking at Volume and Standard Deviation levels.

The use of the Normal Standard Distribution curve can result in Negative Value Levels ( ie. for the Lower Standard Deviation levels ).

Volume can only be a Positive Value, hence I use a LOG Normal Distribution of Volume to assess the Standard Deviation levels. This gives levels which are always Positive. The following Amibroker Code shows what I am referring to.
I believe TG uses this approach also. Is this what you do also ?

All feedback on this line of thinking is most welcome.

Regards,
V

LBP = Param("Look Back", 30, 0, 100,1 ); Mean = MA(ln(V),LBP); StD = StDev(ln(V),LBP); xp3 = exp(mean + 3*std); xp2 = exp(mean + 2*std); xp1 = exp(mean + 1*std); xm = exp(mean); xn1 = exp(mean - 1*std); xn2 = exp(mean - 2*std); Plot(xp3,"", 1,1|4096); Plot(xp2,"", 1,1|4096); Plot(xp1,"", 1,1|4096); Plot(xm, "", 29,1|4096); Plot(xn1,"", 1,1|4096); Plot(xn2,"", 1,1|4096); Clr = IIf(V>Ref(V,-1),29,32); Plot(V,"",Clr,6);

[RAVIN]

	Quote:	TinGull
	The last trade ended up going for 60 points, too...I didn't have time to sit and manage it though. But, VSA came to save the day once again!

DAX Trading was exciting too. CLASSIC ENTRY ON THE CURSOR, LOW VOL TEST, PURE VSA. Exited rather early at S1, expecting it to act as resistance, however there was an opportunity to jump aboard again, but was ruminating over how did I miss that 100pt drop

Looking at your YM chart and the second trade, there was a similar opportunity to go long on the classic down bar test on low vol.

[RAVIN]

Correction TG,

After your 1st trade (which was stopped out), there was a classic test on low vol, ideal signal to go long as per VSA

[rayl54292]

	Quote:	VictorH
	Hello rayl54292, This is my first post on this forum. I was interested in the chart in your last post [3] and your comment about “input and brutal honesty� This is my analysis of VSA in terms of looking at Volume and Standard Deviation levels. The use of the Normal Standard Distribution curve can result in Negative Value Levels ( ie. for the Lower Standard Deviation levels ). Volume can only be a Positive Value, hence I use a LOG Normal Distribution of Volume to assess the Standard Deviation levels. This gives levels which are always Positive. The following Amibroker Code shows what I am referring to. I believe TG uses this approach also. Is this what you do also ? All feedback on this line of thinking is most welcome. Regards, V LBP = Param("Look Back", 30, 0, 100,1 ); Mean = MA(ln(V),LBP); StD = StDev(ln(V),LBP); xp3 = exp(mean + 3*std); xp2 = exp(mean + 2*std); xp1 = exp(mean + 1*std); xm = exp(mean); xn1 = exp(mean - 1*std); xn2 = exp(mean - 2*std); Plot(xp3,"", 1,1|4096); Plot(xp2,"", 1,1|4096); Plot(xp1,"", 1,1|4096); Plot(xm, "", 29,1|4096); Plot(xn1,"", 1,1|4096); Plot(xn2,"", 1,1|4096); Clr = IIf(V>Ref(V,-1),29,32); Plot(V,"",Clr,6);

Victor H,

Fished the Std info from a PivotProfiler post http://www.traderslaboratory.com/for...5-post143.html :

"We also look at the individual volume histogram by itself. Here we are looking at actual size (amount does not matter). A 30 period moving average can be used to determine if volume is greater than average or less than average. If you put a 2.0 standard deviation of that 30 average on the chart as well, then you could call all volume histogram bars greater than that line high. You could also place a 3.0 standard deviation of the average on the chart and call all volume histogram bars greater than that Ultra High Volume. This is basically what TG does."

tradestation code:

var: avg(0), sDev(0);

avg=average(v,30);
sDev=standardDev(v,30,1);

plot1(avg+3*sDev,"+3");
plot2(avg+2*sDev,"+2");

Not a programer, got code from tradestation fourm.

R

[VictorH]

Hello rayl54292

Thanks for your reply. I was aware of Pivotprofiler’s calculations for St Dev levels. I think the present TS Calculation (as you have shown) is flawed when you apply the basic principles of statistics to Volume.

If you apply the TS coding and plot the following lines you will get positive and negative values.

Plot3(avg-2*sDev,"-2");
Plot4(avg-3*sDev,"-3");

Now, do you still have faith in the following 2 lines ???

Plot1(avg+3*sDev,"+3");
Plot2(avg+2*sDev,"+2");

Calculations based on Lognormal do not create negative values.

In the overall scheme of things, the actual calculation is probably trivial and of minor consequence to trading outcomes. “High Volume� is still “High Volume�.

Keep up the excellent posts on VSA.

Regards,
V

[TinGull]

[TinGull]

[rayl54292]

[quote=VictorH;20736]Hello rayl54292

Thanks for your reply. I was aware of Pivotprofiler’s calculations for St Dev levels. I think the present TS Calculation (as you have shown) is flawed when you apply the basic principles of statistics to Volume.

If you apply the TS coding and plot the following lines you will get positive and negative values.

Plot3(avg-2*sDev,"-2");
Plot4(avg-3*sDev,"-3");

Now, do you still have faith in the following 2 lines ???

Plot1(avg+3*sDev,"+3");
Plot2(avg+2*sDev,"+2");

Calculations based on Lognormal do not create negative values.

In the overall scheme of things, the actual calculation is probably trivial and of minor consequence to trading outcomes. “High Volume� is still “High Volume�.

Keep up the excellent posts on VSA.

Regards,
V[/QUOTE

V,

Thank you for the feedback, appreciate that.


R.

[rayl54292]

Hello.

Just wanted to share some intraday VSA/harmonics combinations.

Attached is 15 min IWM chart over past few days. points X,A,B,C,D form a bullish gartley pattern ( http://www.harmonictrader.com/price_patternsgartley.htm ). Test bar (I believe it to be a test) confirms the pattern.

I trade the ER2 of an IWM chart, less noise IMHO, and do not hold over the weekend, thus my entry was the second signal, confirmation (result) after effort to rise.

R

[EyeQue]

Hi TinGull and others:

Excellent Threads. Because of these threads, I am currently analyzing information provided in these threads and the VSA book. After I get comfortable, I will start posting.

Regards,

EyeQue