Hi,

how can we calculate slope of 10 periods moving average?

A net slope from the previous bar to the current bar could be written as follows.

AVGC10 - AVGC10.1

While the percent slope from one bar to the next can be written as follows.

100 * (AVGC10 / AVGC10.1 - 1)

You could adjust this to say 10 bars by adjusting the bars ago parameter. So the net slope would be:

AVGC10 - AVGC10.10

And the percent slope would be:

100 * (AVGC10 / AVGC10.1 - 1)

Once you go to looking at multiple bar slopes, you could calculate a linear regression slope as well.

(4.5 * AVGC10 + 3.5 * AVGC10.1 + 2.5 * AVGC10.2 + 1.5 * AVGC10.3 + .5 * AVGC10.4 - .5 * AVGC10.5 - 1.5 * AVGC10.6 - 2.5 * AVGC10.7 - 3.5 * AVGC10.8 - 4.5 * AVGC10.9) / 82.5

And express this a percentage of the moving average (or some other basis)

100 * (4.5 * AVGC10 + 3.5 * AVGC10.1 + 2.5 * AVGC10.2 + 1.5 * AVGC10.3 + .5 * AVGC10.4 - .5 * AVGC10.5 - 1.5 * AVGC10.6 - 2.5 * AVGC10.7 - 3.5 * AVGC10.8 - 4.5 * AVGC10.9) / 82.5 / AVGC10

Using Linear Regression vs Classical Peaks/Valleys for Divergence Analysis

Hi,

i have just copy and paste 

100 * (4.5 * AVGC10 + 3.5 * AVGC10.1 + 2.5 * AVGC10.2 + 1.5 * AVGC10.3 + .5 * AVGC10.4 - .5 * AVGC10.5 - 1.5 * AVGC10.6 - 2.5 * AVGC10.7 - 3.5 * AVGC10.8 - 4.5 * AVGC10.9) / 82.5 / AVGC10

into tc2000 and got message 'Error in formula Argument missing near *'.

Please advise how to solve this issue.

BR

My best guess would be that it is the use of a . for the decimal marker if your computer is set to use the , as the decimal marker. Here is a re-write which avoids the use of decimal markers altogether.

100 * (9 * (AVGC10 - AVGC10.9) + 7 * (AVGC10.1 - AVGC10.8) + 5 * (AVGC10.2 - AVGC10.7) + 3 * (AVGC10.3 - AVGC10.6) + AVGC10.4 - AVGC10.5) / 165 / AVGC10

thank you!

You're welcome.