Market Profile Graphic

[wildfiction]

In the attached image of the MP chart you can see that the developing VAH and VAL have been set at 58970 and 58690 respectfully.

From 58920 (which has 4 TPOs) we move up 5 ticks to get to the 58970 level. At each price there are 3 TPOs.

From 58710 (which has 4 TPOs) we move down 2 ticks to get to the 58690 level. At each price there are 3 TPOs.

So why did the MP chart decide to cover 5 ticks (of 3 TPOs) each to the upside and 2 ticks to the downside when calculating the developing VAH and VAL?

Why did it not balance it and count 4 to one side and 3 to the other side?

This does not seem like a correct calculation of the developing VAH/VAL to me?

[DionLoy]

Hm, you can't calculate the Value Area based on visual 'balance' of the extremities.

(well, perhaps you could if the middle part was all equivalent, but in your screenshot, the middle portion is not completely symmetrical).

The value area is 68% of the TPOs, but using the Point of Control as the center point. Visually, it seems that the point of control is a bit lower than center of the range, hence needing more price points above to balance out the two standard deviations necessary.

[wildfiction]

Originally posted by DionLoy
Hm, you can't calculate the Value Area based on visual 'balance' of the extremities.

(well, perhaps you could if the middle part was all equivalent, but in your screenshot, the middle portion is not completely symmetrical).

The value area is 68% of the TPOs, but using the Point of Control as the center point. Visually, it seems that the point of control is a bit lower than center of the range, hence needing more price points above to balance out the two standard deviations necessary.

Thanks for the reply DionLoy,

So how does Market Profile 'arbitrate' between 2 prices which have equal TPOs?

Say there was perfex symetry as in this example:

120 X
119 XX
118 XX
117 XX
116 XXX
115 XX
114 XX
113 XX
112 X

The POC is easy to calculate. How does the MP graphic decide which other TPOs to use?

Thanks

[DionLoy]

First, we'll sum up the total number of TPOs in the chart, and then calculate 68% of that total, call this X.

Then, add to it the TPOs in the line ABOVE the POC.

If it's still less than X, add to it the TPOs in the line BELOW the POC.

If it's still less than X, add to it the TPOs TWO lines ABOVE the POC.

etc ...

until your Value Area encompasses at least X.

[wildfiction]

Thanks DionLoy,

From what you have just explained though that algorithm sounds symetrical to me (and logical and sound) but doesn't explain the lack of symetry in the graphic that I posted?

Rather than waste your time here, can you point me to where I can find some documentation on the algorithm that is used in the eSignal Market Profile graphic? That way I can work through the documentation and try and work out my lack of understanding in how it is calculated in eSignal.

A couple of sources that I have found so far are:
1. Appendix 1 in James Dalton's Mind over Markets
2. CBOT Market Profile downloadable PDF

but neither of those algorithms seem to match what is produced in eSignal's Market Profile graphic.

Thanks again.

[DionLoy]

The algorithmn that eSignal uses can be found in Steidlmayer on Markets in the chapter that Steve Hawkins wrote (both of them consulted with us on the creation of our MarketProfile charts).

Actually looking at the code and book now, it is slightly different than what I described, ie, instead of looking at one line at a time, it looks at two above (Y) and two below (Z), then if Y > Z adds Y to the value area, if Z > Y, adds Z to the value area.

If Y > Z, it continues to search two lines above the last two lines, but keeps the two lines below at their same position, at least until Z > Y, then the reverse happens.

So, since in the chart you posted, the mid-section of the TPOs was NOT symmetric, depending on how the algorithmn plays out, you would get the value area as calculated by eSignal. Especially since the point of control was not dead center in your chart.

Hope that helps, took me a few readings of the chapter to figure out what was going on =).

[wildfiction]

Thanks DionLoy,

That description matches more closely with the CBOT and Dalton algos.

What happens when Z == Y ?

[wildfiction]

From inspection and examining MP charts it looks like it favors the upper pair of TPO lines when Z == Y.

Many thanks for your help DionLoy - much appreciated.

[wildfiction]

Also, I assume that when you calculate 68% of the total TPOs you do not include the TPO count at the POC in the summation that you do?

[DionLoy]

Correct. Glad we're on the same page =).

[wildfiction]

Thanks for your help DionLoy.