Hello, Pajo
I’m using 5pm New York time, which [B]currently[/B] is 2100 GMT (beginning November 6, 5pm New York time will be 2200 GMT).
No weighting.
I am presently experimenting with a linear regression line drawn on the Daily ATR(2). That is, my trading platform will draw a graph of the daily average true range (Daily ATR) averaged over any number of days that I specify (from 2 to 200). I’m using the shortest average available, the Daily ATR(2). Then I’m overlaying a linear regression line, and simply reading the pip-level of the linear regression line where it intersects the centerline of each daily candle.
You may have noticed the linearity of the ADR figures that I posted. Until the linear regression line changes, those figures will continue to be linear in their relation to one another.
Over the very recent past, the Daily ATR(2) exhibited lower highs and lower lows between 8/1/11 and 9/2/11. Then, it exhibited higher highs and higher lows from 9/2/11 through 10/7/11. From 10/7/11 through yesterday, the Daily ATR(2) has once again exhibited lower highs and lower lows.
Accordingly, the linear regression line superimposed on the Daily ATR(2) is a series of three line segments, end to end. The first (8/1/11 - 9/2/11) slopes downward; the second (9/2/11 - 10/7/11) slopes upward; and the third one (10/7/11 - present) slopes downward.
The linear regression line indicates an ADR of 126 pips for today (5pm Wednesday - 5pm Thursday).
Edit: I have corrected the ADR for today.