Bruce, can you suggest how one can plot three sets of lines around a moving average: at one, two, and three ATRs above and below EMA 13?

This idea is from Kerry Lovvorn and is discussed in  "The New Trading for a Living" by Alexander Elder on pages 93 and 94 and 221.

Add three Custom PCF Channel Indicators to the chart with the following settings in TC2000 v17.

The Width Multiplier needs to be different for each of the three Custom PCF Channel Indicators. Set it to 1.00 for one of them, 2.00 for one of them, and 3.00 for one of them.

Thanks for the instructions, Bruce.

However, I've tried the Custom PCF Indicator to no avail. 

Please be more explicit about how to start with the "Custom PCF CHANNEL Indicator".

Maybe you mean Envelop Channel which allows for Width...

It is pretty much the same as adding any other indicator. You just have to choose the correct indicator when adding it to the chart (so select Custom PCF Channel Indicator instead of Custom PCF Indicator, Custom PCF Cumulative Indicator, or Custom PCF % True Indicator).

I've clicked on Ema13, then "Add Plot Here", then search for "Custom PCF Channel Indicator", but it is not on the list.

I am using V17 ending in .28927.

It appears when I click on the stock Symbol, but not when I click on Ema13. 

Aren't we just talking about 3 Keltner Channels all wrapped around the same EMA(13)? Just set the EMA mulitpliers to 1, 2, and 3 respectively?

Now it's working per your instructions.  Thanks, again.

Just click on Add Plot to Chart (the green + in the upper left corner of the chart) instead of clicking on your exponential moving average and selecting Add Plot Here.

Or if you really want to use Add Plot Here instead of Add Plot to Chart, click on Price instead of the exponential moving average.

By the way, I see you've picked up your library copy ;)

The difference between this and Keltner Channels would be the centerline. This is using a plain old 13 period exponential moving average of price as the centerline while Keltner Channels would be using a 13 period exponential moving average of typical price ((H + L + C) / 3).

Here's the 3-ATR channel in the book that traderlady is referring to. Think 3 Keltner Channels wrapped around the same 13-period EMA centerline will do...

OK, try an experiment.

Add the Custom PCF Indicator described above to the chart (you can add it three times or just once for demonstration purposes).

Then add Keltner Channels to the chart (and I mean to price, not to the EMA13). You can again do this three times or just once for demonstration purposes.

Make sure the settings match.

The indicators are going to be similar, but not identical for the same reasons as when comparing the Keltner Channel centerline to the Bollinger Bands centerline.

You can add a 13 period exponential moving average to the charts to see which of the indicators has a centerline which matches (spoiler: it should be the Custom PCF Channel Indicator).

Now, actually add the Keltner Channels to the 13 period exponential moving average if you want. The other two were similar, but this one ends up being very different.

No need for an experiment, Bruce. I was in the middle of posting that chart while you were posting your 12:04:12pm response :)

First, I meant ATR multiplier instead of EMA multiplier, but I'm sure you got that.

I'd pulled up my chart from the whole parabolic discussion we've been having. I didn't see typical price as a choice for Average Type.

Then I checked the Help File. I totally forgot Keltner Channels used typical price. And I have built Custom PCF Channel Indicators for ATR channels in the past for this very reason. I just spaced out for a minute there.

Thanks for the reminder, Bruce! I'm not crazy (just seem that way online some time) my mom had me tested :)

We now return to our regular programming.

Of course, that now begs the quesion whether an ATR channel or a Keltner channel is more appropriate for the parabolic scan and the chart. 

Bruce, I plotted a Keltner Channel on my chart, but admit since you're the Jedi Master I didn't take a close look at your PCF code.

Can I assume your parabolic PCFs are still good as they are? I'm thinking yes - just want to make sure.

So, what I'm thinking is I just need to replace the Keltner channel on my chart with a Custom PCF Channel Indicator to be 100% kosher. In that one example it didn't seem to make a difference but you always want to have as accurate info as possible when trading with real $$.

The formulas in Parabolic and parabolic-ish uptrends used a moving average of the close for the centerline and not a moving average of typical price (so it didn't use Keltner Channels).

Cool, thanks Bruce. Good to know that my one senior moment was an isolated incident and not the beginning of a pattern. Only had three cups of coffee today. Usual dosage.

I'll replace the Keltner Channel on my chart with a 3-ATR Custom PCF Channel Indicator. That should make what I'm seeing on the chart match what I'm seeing in the PCF/EasyScan results. Don't really want another senior moment. :)

And I'm wrong....

The formula I posted after reading through your original post used a moving average of the close as the centerline.

But the formulas I posted after you posted a picture using Keltner Channels use a moving average of typical price.

So Bruce, how much coffee have you had today? ;)

Kidding. My bad. That thread now has a lot of crap in it, but the concept and I believe the final version of the parabolic PCF is worth it.

I know I kind of highjacked this thread to talk about the other one, but...

If you haven't already done so (you type faster than I do) can you post the parabolic PCF(s) with a centerline using closes instead of typical price? Maybe at the very end of that thread where it'll be easy to find. Other folks might actually find the PCFs useful, if they can make it through all the other stuff. LOL.

I don't know Elder personally, but I know enough that he picked up the idea of a 3-ATR channel from Kerry Lovvorn, who traderlady mentioned earlier today. And he loves his AutoEnvelope Channel. I'd bet your life that he's not into Keltner Channels, so our parabolic PCF shouldn't use typical price if we want to be completely accurate.

Hello Bruce:

I'm following along so far, but am interested what it would look like to use a separate period for the ATR than the EMA. 

For example, I would like a pcf that will give me this:  20ema + 2(10ATR) and 20ema-2(10ATR) as one of the three channels.

thanks

nimble8

You may want to look at the based formulas for Keltner Channels in the following help file.

It has all of the calculations in a single XAVG() function. We can divide that into two XAVG() functions. One for the centerline and one for the ATR or we can just use the built in ATR syntax (although that would result in a simple moving average instead of an exponential moving average for the ATR portion).

I am guessing you don't want to use typical price for the centerline based on the phrasing of your question.

For the exponential ATR channel width:

Upper Channel:

XAVGC20 + 2 * XAVG(ATR, 10)

Centerline:

XAVGC20

Lower Channel:

XAVGC20 - 2 * XAVG(ATR, 10)

For the simple ATR channel width:

Upper Channel

XAVGC20 + 2 * ATR10

Centerline:

XAVGC20

Lower Channel:

XAVGC20 - 2 * ATR10