Help with Keltner Bands

[jthursby]

I need help with matching up the standard Keltner Bands (Keltner.efs) in eSignal to those produced in the Ensign charting software. Within Ensign, you have the ability to introduce a multiplier - something not available in the standard eSignal study. The exact parameters that I'm seeking to match are:

Multiplier of 2
Ave (period) of 9

Thanks!!!

[ACM]

ChiSales
You should be able to find what you are looking for here
Alex

[DuaneG]

Hi jthursby,

You may also want to read through this article from the EFS Knowledge Base. It contains the raw code and the developer comments regarding the study that Alexis referenced.

[jthursby]

I've looked at the Ensign Keltner Band that I am trying to replicate, and I think that it is based off of a Simple Moving Average. . . versus the Extended Moving Average that is used in all of the posted modified Keltner Bands. I think with that change, it might make the difference. Does
anyone supply a version that has that?

Thanks!

[ACM]

ChiSales
You can find a version of the Keltner channel that uses simple moving averages here. This one replicates the version available in TradeStation.
Available here instead is a variant of one of the Keltner studies from the link in my prior message. This one replaces the exponential moving average with a simple moving average and adds alerts (which can be easily removed)
FYI I found both of these by running a Search for Keltner.
Alex

[eSignal_JasonK]

Hello Jonathan,

Did either of the studies Alex pointed you to match your ensign study?

[matt hellrich]

Hi Alex,

let me butt in for a second in regards to searches.
I am not too good at finding anything with esignal so maybe I go the wrong way about it.
If I want to find a formula, I normally try the file sharing search which typically finds about zero.
So I guess searching esignal central is the way to go.
Why don't I ever find anything under file sharing ?
Thanks as usual for all your assistance.

Matt

Originally posted by Alexis C. Montenegro
ChiSales
You can find a version of the Keltner channel that uses simple moving averages here. This one replicates the version available in TradeStation.
Available here instead is a variant of one of the Keltner studies included in the link in my prior message. This one replaces the exponential moving average with a simple moving average and adds alerts (which can be easily removed)
FYI I found both of these by running a Search for Keltner.
Alex

[jthursby]

the most recent version that was recommended is very -very close, but still off by a few ticks on the upper and lower bands. The center moving average line is spot on. I haven't yet, but my next step is to try a different multiplier than 2. . . perhaps 2.1, or something similar, etc. . .

to give you a point of reference, sample data points that I'm trying to match would be in Ensign:

On a weekly December Ten Year Notes (TY Z4) chart, the three bands for some random dates are:

10/22/04
114-318 Upper
112-170 Middle (simple MA)
110-022 Lower

9/17/04
113-269 Upper
111-170 Middle (simple MA)
108-022 Lower

8/4/04
111-132 Upper
108-207 Middle (simple MA)
105-283 Lower

thanks,
Jonathan

[eSignal_JasonK]

Hello Jonathan,

Ok, so all we need to find out now is what formula the bands are based on. I can think of three options. There's probably more, but it's likely to be one of these three.

1. Average True Range over a specified period * multiplier
2. Average Range over a specified period * multiplier
3. Fixed percentage from basis line

If you could post an image of the ensign chart, we should be able to figure it out.

[eSignal_JasonK]

Hello Jonathan,

Thanks for sending me the ensign images. Looks like the attached formula, which is Alex's version he made for the trade station reproduction, does the trick. The key was the price source. I noticed in the parameters window you sent me, the price source was set to HLC/3. I saved the defaults for you 9, 2 with HLC/3 as the default price source in the attached. You should have an exact match now. By the way, ensign bases their bands on the same formula as TS, which is the difference of MA of High - MA of Low.

ensign

[jthursby]

Jason and Alex. . . thank you!

It looks like you've found the winning combination. Outstanding!

Jonathan

[jthursby]

All. . . everything is now moving forward very well with the latest iteration of the Keltner Bands. The only aspects that are on the wish list are some cosmetic tweaks. 2 things were trying to do:

1. Give the chart a black background - when the colors are changed, the O, H, L, L values can not be seen. The background behind the values are the same color as the text and it is not possible to have them set differently.

2. Change the colors of the Upper and Lower Bands. In changing the properties, there is only one color field available. . . which turns all 3 bands into same color. Ideally, each can be altered uniquely. . . or at least the the middle MA line one color and the Upper/Lower bands another.

Any help out there in modifying the last posted version of the Keltner Bands would be greatly appreciated.

thanks,
Jonathan

[ACM]

Jonathan
The attached revision of the efs adds the option to color the Keltner basis and band lines through Edit Studies
Alex

[ACM]

Jonathan
With regards to the color of the OHLC values right click the Cursor Window, click the Font button and select a different color
Alex

[jthursby]

Got it!!! Everything is set cosmetically now. Changing the colors in the dataview box is a little querky. . . but I found a combination that works. I can't thank you enough for all your help and prompt turn around on postings!!

Jonathan