[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

FW: Why futures truth doesn't care about stealing your system


  • To: "List, Omega" <omega-list@xxxxxxxxxx>
  • Subject: FW: Why futures truth doesn't care about stealing your system
  • From: "Brian" <blink64@xxxxxxxx>
  • Date: Sat, 31 Mar 2001 21:01:50 -0800

PureBytes Links

Trading Reference Links


For all the power users out there.

http://www.mathworks.com/products/mathlibrary/


-----Original Message-----
From: Brian [mailto:blink64@xxxxxxxx]
Sent: Saturday, March 31, 2001 8:52 PM
To: List, Omega
Subject: RE: Why futures truth doesn't care about stealing your system


" basically you have to differentiate between a
> - trading platform and
> - research platform"

Agreed.  TS wasn't designed to be pure math research platform.  You can't
criticize TS for being an inadequate research platform because it wasn't
designed to be Mathlab.  I feel that for the price, it does a lot and offers
most people an endless amount of flexibilty.  You can't backtest in Mathlab
can you?  You can't auto monitor positions in ML right?  TS was designed to
be a trading platform that does all the nec things fairly well for a
reasonable price.  Most people wouldn't know what to do with MathLab.

How would you port MatLab and Mathcad to TS?  Develop an api interface
between TS and ML?

> -----Original Message-----
> From: Bilo Selhi [mailto:citadel@xxxxxxxxxxxx]
> Sent: Saturday, March 31, 2001 5:41 PM
> To: orcas@xxxxxxxx; List, Omega
> Subject: Re: Why futures truth doesn't care about stealing your system
>
>
> Matlab and MathCad are serious math and engineering
> apps. you can't buy an ocx that can replace those.
> matlab language is similar to C but specialized.
> you can't compare C dev tool with math platforms.
> Matlab or Splus are interactive math platforms. you use it
> for research. once you are done with your research then
> you can hard code your code in C for efficiency.
> basically you have to differentiate between a
> - trading platform and
> - research platform
> tradestation is mix of both but it's no match to number crunching
> ability of Matlab or Splus. but as same time those are
> not adapted to trading.
> if you put both together you get the best of both worlds.
> and my personal experience is that tradestation is on it's way
> of  becoming a decent trading platform on which you can depend.
> although it doing it very slow. the drawback there is that
> tradestation is not yet mission critical application and not yet
> fast and dependable enough to be used for high end heavy load
> professional trading. it's getting there but slow.
> bilo.
> ps. i Trad was smart enough they port TS both to matlab and mathcad
> and i am sure they'd start picking up some institutional and professional
> clientele
> FYU, both Matlab and Splus are ported to Bloomberg. this
> combination is the
> standard research platform for top large research and trading
> firms and high
> tech
> financial engineering firms. for them tradestation is somewhat an
> interesting toy...
>
> ----- Original Message -----
> From: "Brian" <blink64@xxxxxxxx>
> To: "List, Omega" <omega-list@xxxxxxxxxx>
> Sent: Saturday, March 31, 2001 4:16 PM
> Subject: RE: Why futures truth doesn't care about stealing your system
>
>
> > "TS is a toy.  We need serious programs that are capable
> > of serious number crunching like what MatLab and MathCad can do."
> >
> > Buy a C++ OCX that does everything (or at least most) MatLab and MathCad
> can
> > do, insert it into your DLL that TS calls and what do you have?
>  You have
> a
> > trading platform that really can do it all.  I think C++ is just as
> flexible
> > if not more than MatLab or MathCad and it's better becaue it
> can work real
> > time.  Aren't MathCad and ML both stand alone apps?  Learning C++ (or
> > powerbasic!) is about as complicated as learning to use MatLab
> or MathCad
> > effectively.  I like that approach better over something like
> MathCad and
> > that's why I'd still choose Tradestation.
> >
> > > -----Original Message-----
> > > From: Mark Cerar [mailto:marQc@xxxxxxxxxxx]
> > > Sent: Friday, March 30, 2001 8:38 PM
> > > To: Lawrence Chan; omega-list@xxxxxxxxxx
> > > Subject: RE: Why futures truth doesn't care about stealing your system
> > >
> > >
> > > Bang on Lawrence!  TS is a toy.  We need serious programs that are
> capable
> > > of serious number crunching like what MatLab and MathCad can do.
> > > Engineers,
> > > statisticians and scientists would not use crap like TS and in
> > > many ways the
> > > demands of real-time financial software are even greater than the more

> > > "static" data they frequently work on.  If OR continues on its present
> > > course, TS in all its permutations and combinations will NEVER,
> > > EVER become
> > > the de facto trading-software standard of the future.  In fact,
> > > it's fate is
> > > being sealed as we speak.  I hope you are all preparing.  The end is
> nigh.
> > >
> > > -----Original Message-----
> > > From: Lawrence Chan [mailto:stnahc@xxxxxxxx]
> > > Sent: March 30, 2001 9:50 AM
> > > To: brian; omega-list@xxxxxxxxxx
> > > Subject: Re: Why futures truth doesn't care about stealing your system
> > >
> > >
> > >
> > > foucs on money management is a waste of time -
> > > IF the base model does not have the necessary
> > > "edge".
> > >
> > > And in fact, there are new models emerged over the
> > > past few years not available to the public that
> > > perform very well during both 1999 and 2000, for 2001
> > > we have to see.
> > >
> > > Excalibur cannot test these new models as the design
> > > of Excalibur just cannot handle it, period.
> > >
> > > The good news is that nor TS can do it at all :)
> > >
> > > -Lawrence
> > > Lawrence Chan                http://www.tickquest.com
> > > Home of trading tools NeoBreadth and NeoTicker series
> > >
> >
>
>