The common wisdom seems to be that there isn't any discernible difference between time frames, but the information tells another story. Back in January I did. The hyperlink to the whole set of study notes is here:Quote:
Originally Posted by ;
http://market-geeks.com/research/archive2008_01_17.doc
But the advice pertinent to this discussion is just on pages 5, 11, and 13, so I'll just paste it here.
If you divide the total range of any arrangement of pubs from the minumum cumulative distance that price has to travel in each one the pubs TR/(avg. Bar height x #bars) you get a measure of just how much of the price action given to the trend and just how much of the price action was dropped to chop. The main point is that the trendiness drops by orders of magnitude as we fall down in time frames.
Here is the excerpts in the notes:
The analysis of the daily statistics is on page 5:
You will find 1358 bars using an average range of 112 pips each. This means there was a motion of 151,520 pips during the period. The total range over the period was from 124.12 into 168.96 or 4484 pips. This range is 3% of their movement, meaning that 97% of the motion was dropped to chop. The bar height is 2.5% of the total range.
From page 11:
Qualities of this 10-minute info:
Bars: 7300
Typical bar range: 22 pips
Total motion: 159,282 pips
Total Range: 909 pips
Trend variable: 0.6%
Noise: 2.4%
From page 13:
Below are the Features of this 10-second bar information:
N 16970
H 5.7
NH 96712
R 136
Trend 0.1%
Noise 4%
H(max) 20