|
PureBytes Links
Trading Reference Links
|
Thats exactly the problem, Gary. While the indicators get enough data to
spin up, I get
the reading slightly off as compared to continuous contract. Also, it gives
different figures
if I want to see the P/L in %age rather than dollars because the whole data
series is adjusted.
This will possibly affect the drawdown %age figure as well but I dont know
for sure.
So I am trying to find a method which tests on real data & accounts for
rollovers. Say it only trades
when the contract has enough data for the indicators also if I test the
individual contracts in succesion,
to export them into excel, I don't get the duplicate trades. Also I need to
find a way to tess TS to exit a
trade at rollover, & open a position in the front-month contract
Reagrds
HT
----- Original Message -----
From: "Gary Fritz" <fritz@xxxxxxxx>
To: "HT" <thakral@xxxxxxxxxxx>
Sent: Wednesday, May 01, 2002 9:31 PM
Subject: Re: Roll-Over
> > I am presently using point-based back-adjusted data for testing,
> > like most of us do. It certainly doesn't give the same figures as
> > the real contract data with rollovers.
>
> If not, then you should look at your systems. Other than at the very
> start of a contract -- when front-contract-only charts give WRONG
> answers because of insufficient data to "spin up" the indicator
> calculations -- my systems produce nearly 100% identical trades on
> front-contract charts or continuous charts.
>
> Gary
>
>
|