|
PureBytes Links
Trading Reference Links
|
Joe
My response was tongue-in-cheek. I'm not defending errors in the MetaStock
system tester, and I wouldn't expect any new release to eliminate all
existing problems. I guess that's life.
Roy
> Thanks for the hint. I'm thinking of upgrading to v8.01 but that is
> not the issue with what I'm talking about here. I've been a long time
> user of Metastock since the real-time DOS version,
> I think it was 4.0 or something. I've been using 7.0 pro in the past
> for system development and testing. Late last year I purchased APS
> from tradingpatterns.com. I was skeptical in the beginning about
> getting another program but that specific one helped me to realize a
> few things about testing systems. When comparing the back testing
> results of APS and Metastock for the same systems they were different
> and that made me very curious. I then went manually, trade-by-trade
> over a number of system testing results and I realized that Metastock
> testing was wrong. The problem is that if a system exited an open
> position on the open or during the day and the same signal generated
> a new position on the close of the same day or the open of next day,
> that specific trade would not show up in the testing. As a result,
> the performance of the system was misleading or even wrong, depending
> on how many trades were skipped by the tester. But here comes another
> problem: If one uses a time shift as suggested to fix the tester
> problem then the same formula in the explorer misses the signal.
>
> I feel very uncomfortable that years of work with Metastock testing
> systems involved such a "limitation". Even if this is solved in v8.01
> that's of no cancellation for all the wrong testing results and the
> missed trades.
>
> Joe
>
>
>
>
>
>
>
>
> --- In Metastockusers@xxxxxxxxxxxxxxx, "Roy Larsen" <rlarsen@xxxx>
> wrote:
> > Hi Joe
> >
> > I think the Equis "solution" is v8.01.
> >
> > Roy
> >
> > ----- Original Message -----
> > From: "equtrader" <equtrader@xxxx>
> > To: <DEFANGED_Metastockusers@xxxxxxxxxxxxxxx>
> > Sent: Tuesday, May 20, 2003 8:50 PM
> > Subject: [Metastockusers] Re: Tester question
> >
> >
> > > Hi Roy,
> > >
> > > Thanks a lot. I checked that and it works fine. But here is the
> > > question: Now, using your shifting, the equity curve is correct
> but
> > > the signals in explorer are received one day after they are
> > > generated. Because if 0 bar is today then the signal on the open
> will
> > > be generated one day late. I just run the tester and an
> exploration
> > > for a date I know the signal was generated. Your method gave the
> > > correct backtest but a signal was not generated for next open. In
> > > other words, you need a different formula for the tested and the
> > > explorer for the same system to produce compatible results. Beyond
> > > that, you may not even be getting the right signal timming if your
> > > are using the formula that's good for the tester only.
> > >
> > > We got a serious problem here, I mean Metastock's got one.
> > >
> > > Joe
> > >
> > >
> > >
> > > --- In Metastockusers@xxxxxxxxxxxxxxx, "Roy Larsen" <rlarsen@xxxx>
> > > wrote:
> > > > I have 7.03 EOD and this certainly seems to be the way it works.
> > > >
> > > > I think a work-around for this would be to change your entry
> code to
> > > > Ref("entry code",-1) and at the same time change the System
> Tester
> > > to zero
> > > > delay. the effect of the two changes is that the timing of your
> > > signal stays
> > > > the same but the System Tester doesn't not assume an overlap and
> > > inhibit the
> > > > trade entry as a result. I have not tested this recently but I'm
> > > aware that
> > > > the treatment of delays is one of the reasons my Trade Equity
> can
> > > produce
> > > > different results.
> > > >
> > > > Roy
> > > >
> > > >
> > > > ----- Original Message -----
> > > > From: "equtrader" <equtrader@xxxx>
> > > > To: <DEFANGED_Metastockusers@xxxxxxxxxxxxxxx>
> > > > Sent: Tuesday, May 20, 2003 8:37 AM
> > > > Subject: [Metastockusers] Tester question
> > > >
> > > >
> > > > > I use 7.0 Pro.
> > > > >
> > > > > I have this simple formula in the tester. I noticed that when
> an
> > > open
> > > > > position exits and a signal is generated on the same day close
> > > with
> > > > > an entry at the open of the next day, the tester does not take
> > > that
> > > > > position. Is this the way the Metastock tester works?
> > > > >
> > > > > Thank you
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > To unsubscribe from this group, send an email to:
> > > > > Metastockusers-unsubscribe@xxxxxxxxxxx
> > > > >
> > > > >
> > > > >
> > > > > Your use of Yahoo! Groups is subject to
> > > http://docs.yahoo.com/info/terms/
> > > > >
> > > > >
> > > > >
> > > > >
> > >
> > >
> > >
> > > To unsubscribe from this group, send an email to:
> > > Metastockusers-unsubscribe@xxxxxxxxxxx
> > >
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
> > >
> > >
> > >
> > >
>
>
>
> To unsubscribe from this group, send an email to:
> Metastockusers-unsubscribe@xxxxxxxxxxx
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
>
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Get A Free Psychic Reading! Your Online Answer To Life's Important Questions.
http://us.click.yahoo.com/Lj3uPC/Me7FAA/CNxFAA/zMEolB/TM
---------------------------------------------------------------------~->
To unsubscribe from this group, send an email to:
Metastockusers-unsubscribe@xxxxxxxxxxx
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
|