|
PureBytes Links
Trading Reference Links
|
My understanding is that TS/SC can be programmed to start trading on a
certain day, and stop trading on another certain day thereafter, *for a
particular contract*.
What I mean when I say TS/SC can't do it, is to backtest from contract
to contract without human intervention. Maybe some supercoder has done
it, or can, but I've never heard of it.
Mark Johnson is a big fan of Trading Recipes, which apparently can do
it, but since I have been unable to get detailed info about the program
from RW, I don't know for sure. I'd like to know. MAybe this is why
Mark uses continuous contracts!
Jim Allen
The Omega Man wrote:
>
> Jim wrote:
>
> > Why not leave the data alone, and program into your testing routine
> > whatever rollover scheme you chose to employ...
> > Of course, you can't get TS or SC to do it, but Excel or VB or some
> > such language ought to be able to get it done.
>
> I don't see any reason that TS couldn't do this... There are a bunch of
> functions in the Data Information/General category in TS2k that could handle
> this gracefully.
>
> All the best,
>
> The Omega Man
>
> ----- Original Message -----
> From: Jim Allen <jallen7@xxxxxxxxxxxxx>
> To: Jose Pascual <jpascual@xxxxxxxxxxx>
> Cc: Robert Bianchi <R.Bianchi@xxxxxxxxx>; <omega-list@xxxxxxxxxx>
> Sent: Sunday, June 13, 1999 4:12 PM
> Subject: Re: Continuous vs Perpetual - which one is the best!
>
> > I have a CD-ROM full of daily OHLCVOI data for contracts going back just
> > about all the way to inception. According to what I am reading here, my
> > choices are to either build a psuedo perpetual file which adjusts every
> > price in the file according to some adjustment scheme, or build a file
> > consisting of actual prices adjusted either forward or backward by
> > adding or subtracting the difference between two successive contracts on
> > some arbitrary date known as the "rollover date".
> >
> > Why not leave the data alone, and program into your testing routine
> > whatever rollover scheme you chose to employ. Run your system on March
> > 1970 Corn until the rollover, then May, then July, etc., etc.
> >
> > Figuring this out should be child's play to some of you supercoders, and
> > then you could really test as you trade!
> >
> > Of course, you can't get TS or SC to do it, but Excel or VB or some
> > such language ought to be able to get it done.
> > Jim Allen
> >
> > Jose Pascual wrote:
> > >
> > > I had observed that too that its pretty hard to justify and it depends.
> So better use real futures contract data to trade while using or choosing
> any of the 20,000 plus available securities from NASDAQ, NYSE, etc... to
> backtest. It should work on any or else your system is called "Optimized".
> > >
> > > At 09:21 PM 6/13/99 +1000, Robert Bianchi wrote:
> > > >Q: Backadjusted versus Perpetual - which one is the best ?
> > > >
> > > >A: It depends on your trading system.
> > > >
> >
|