|
PureBytes Links
Trading Reference Links
|
Gene wrote:
>And there's a compelling reason to figure out what TS6 is doing
>differently...
>
>The results are UNIFORMLY better in TS6... if it can be figured out, it can
>be exploited. But gotta be able to backtest it.
That's interesting. I wonder if the fault is in the ROC function or
the way TS6 handles fills versus TS2000i.
Is the source code for RateOfChange the same in both programs? It looks
pretty trivial in 2000i. Perhaps the outputs of RateOfChange could be
printed out and compared?
If you wrote your own RoC function and used it in both software
versions, how do the results compare?
--
,|___ Alex Matulich -- alex@xxxxxxxxxxxxxx
// +__> Director of Research and Development
// \
//___) Unicorn Research Corporation -- http://unicorn.us.com
|