Trading with Market Statistics. IV Standard Deviation

[jperl]

	Quote:	Soultrader
	I have another question regarding the SD. How narrow is narrow? It appears that Jerry would take trades with risk:reward ratios of slightly below 1.0. I personally do not take these type of trades even if Im scalping. If scalping the minimum I aim for is a 1:1.5 risk reward ratio. I understand that these risk:reward thinking is not entirely correct as it is geared for newer traders. However, I just never found myself worth taking a trade with so little reward opportunities.

To restate this question, how large does the SD have to be before you enter a trade?
If you are only using today's distribution, then at the open, the SD is virtually zero. As more price/volume data is added, the SD grows. What's nice about watching this is you can see how the volatility is changing as the day continues. First growing, possibly leveling off, then growing again, then reaching some kind of stasis.
So when should you first consider taking a trade? My rule of thumb is, don't trade until the range of the bars on your chart are less than 1 standard deviation. If the high of a bar touches 1 SD level and the low of the bar touches a second SD level, then the bar range is too large to consider entering a trade on that time frame. You either need to wait until the bar range gets smaller, or go to a faster time frame.

[smodato]

	Quote:	NAVEEVIa
	Can somebody please advise me on this code. I am a newbie to Easylanguage. I am using this code on mullticharts Comments appreciated Naveen vars: vwap(0), pv(0), Totalvolume(0), Barfromstart(0), Squareddeviations(0), Probabilityweighteddeviat ions(0), deviationsum(0), standarddeviation(0); If date > date[1] then begin Barfromstart=0; pv=AvgPrice*volume; Totalvolume=volume; vwap=pv/totalvolume; end else begin Barfromstart=Barfromstart[1]+1; pv=pv[1] + AvgPrice*Volume; Totalvolume=Totalvolume[1] + Volume; vwap=pv/Totalvolume; end; deviationsum=0; for value1= 0 to Barfromstart begin Squareddeviations=Square( vwap-avgprice[value1]); Probabilityweighteddeviat ions=volume[value1]*Squareddeviations/Totalvolume; deviationsum=deviationsum +Probabilityweighteddevia tions; end; standarddeviation=SquareR oot(deviationsum); plot1(vwap); plot2(vwap+standarddeviat ion); plot3(vwap+2*standarddevi ation); plot4(vwap-standarddeviation); plot5(vwap-2*standarddeviation);

Hi,
that code works properly for me, I see we are all worried about VWAP to be plotted in Tradestation; but my real concern is how to plot PVP, if I use AVGprice function I never have "round" prices to work on so how can I say what price showed the maximum volume in the day?
Bye
Smodato

[darthtrader]

hi, check out the thread in Trading Indicators. Blowfish looks like he should have everything pretty much nailed down at some point as far as matching Jerry's videos.

[smodato]

	Quote:	darthtrader
	hi, check out the thread in Trading Indicators. Blowfish looks like he should have everything pretty much nailed down at some point as far as matching Jerry's videos.

Thanks for the suggestion, unfortunately I did not get the PVP code there, furthermore other codes I found are ELD format and using Prosuite 2000i I cannot read them, please see this attachment and maybe someone can read it and post it here in order for me to "transport" it into prosuite 2000i.
Thanks, bye
Smodato

[mmaker]

Howard Arrington just told me he is including SD as a built in study in ensign to be released today in the nov 8th beta.

Looking forward to seeing this.

In his words.....

"VWAP with standard dev band based on price-VWAP is now a built in study in the 11-08 beta to be released later today.
Enjoy. You can do away with the DYO for VWAP now."

[jperl]

	Quote:	mmaker
	Howard Arrington just told me he is including SD as a built in study in ensign to be released today in the nov 8th beta. Looking forward to seeing this. In his words..... "VWAP with standard dev band based on price-VWAP is now a built in study in the 11-08 beta to be released later today. Enjoy. You can do away with the DYO for VWAP now."

Glad to hear it. It was only a matter of time.

[LS_Chad]

Would it not make more sense to convert the VWAP bands into an oscillator. Here is essentially what it would look like:

http://www.charthub.com/images/2007/...Oscillator.png

This way, you have one histogram, instead of 5, 7, 9 lines overlaying your candles. The histogram tells you how many standard deviations price is above or below the VWAP. I'll add this oscillator option to our next release (I/RT 9.0).

[jperl]

	Quote:	LS_Chad
	Would it not make more sense to convert the VWAP bands into an oscillator. Here is essentially what it would look like: http://www.charthub.com/images/2007/...Oscillator.png This way, you have one histogram, instead of 5, 7, 9 lines overlaying your candles. The histogram tells you how many standard deviations price is above or below the VWAP. I'll add this oscillator option to our next release (I/RT 9.0).

That's fine for the forward candles, (the ones all the way to the right). However the oscillator gives the wrong impression about where price has been relative to the standard deviations in the past, due to the fact that the standard deviation lines are renormalizing with each added data point.
For example from 12:20 to 12:50, the lows of all the candles touched the 3rd standard deviation price which kept renormalizing to a lower price.
The oscillator on the other hand gives the impression that none of those candles touched the 3rd standard deviation until about 12:40

[MidKnight]

	Quote:	Soultrader
	Thanks Blowfish, This will indicate that a trader using this methodology must wait until the VWAP, SD, and PVP becomes clear? The SD usually takes all morning before widening up for the Nikkei. This is why I was thinking of applying this technique for the afternoon only. However, I am trying to figure out a way to play the open. For example, if price is below the previous day VWAP and PVP go short, etc.... Im curious to here any comments on such strategy. Perhaps add a few filters like candle patterns, etc...

Hi Soultrader,

I used to use several current day volume based studies on the Nikkei but found that the opening print often ends up distorting the information for most of the morning as you have commented on above. I ended up changing the studies to tick based or time based. Maybe take a look at those for alternatives.

My best regards,
MK

[Kiwi]

Interesting you should say that MK. I have recently adjusted my Volume bars to self-scale for the last 20 bars but I exclude the opening bar and, for SPI, include all bars after the close of the cash market. This makes Wyckoff style volume use much more effective.

How are things going? (pm me )