|
PureBytes Links
Trading Reference Links
|
Riccardo Ronco wrote:
>
> first the user function:
>
> { Function: ChoppyIndex }
> If HighestBar(TrueHigh,30) <= LowestBar(TrueLow,30)
> then ChoppyIndex = 10*(Highest(TrueHigh,30)-Lowest(TrueLow,30))
> else ChoppyIndex = - 10*(Highest(TrueHigh,30)-Lowest(TrueLow,30));
>
> why make things more complicated?
>
> I'd like to see something like this:
>
> ChoppyIndex = 100*(Highest(TrueHigh, Length)/Lowest(TrueLow, Length));
>
> so the indicator is based on a percentage value... and good to compare
> apples with apples and to see if there are any regularities in parameters
> for different commodities.
>
> .........
I like your improvement that makes it applicable across all price
scales. But I still wonder how well it identifies what most people
would consider choppy/trendy. Consider 2 markets both with truehigh at
300 and true low at 100. One has been flailing around with 50 point
daily ranges. The other has been making more or less regular moves up
from 100 to 300. Both would calculate the same amt. of choppiness by
the original and modified indicies.
How about CI = (close - close[length]) /
average(truerange,length)*length ?
The closer this is to 1.0, the trendier the mkt. (This may be similar
or the same as Kaufmans' efficiency ratio?)
(Like ADX and Rsquared this really tells you that the mkt *has been*
choppy/trendy, but what's it going to do in the future???)
Think this would permit the use of one set of parameters for all
commodities?
Conrad Bowers
|