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

Re: Why bad realtime results?



PureBytes Links

Trading Reference Links

Jack:

I've seen the same thing, but the other way around: good real time
trades that disappeared when the "same" historical data was loaded. I don't
use bouncing ticks, and it had nothing to do with trailing stops
etc. I think it will be difficult for someone to tell you why it happened
to you without looking at the specific data and code, but it could be time
related.

I was using TS4/DS/QFeed when it happened here.

BW

----- Original Message -----
From: "Thomas Brun" <tebrun@xxxxxxxx>
To: "Jack Griffin" <jack_2231@xxxxxxxxx>
Cc: <omega-list@xxxxxxxxxx>
Sent: Thursday, August 09, 2001 6:37 AM
Subject: Re: Why bad realtime results?


> I had a similar experience years ago when I took a look at a system
> called broken arrow.  I also was able to create the same symptoms using
> systems on Point and Figure charts.  In my case the problem was to do
> the way Trade Station handles bouncing ticks.  Your problem may be
> similar but I am not familiar with 2000i or Dynastore or quote.com.  My
> guess is that bouncing ticks is your problem.  Sounds exactly similar to
> my experience.  I might add the results of the above systems were
> exceptional but false.
>
> Hope that helps.
>
> Regards,
>
> Tom
>
> Jack Griffin wrote:
>
> > Hi gang,
> >
> > I started running a new intra-day system realtime via
> > Tradestation 2000i + Dynastore + quote.com.  After a
> > several days of running it, I am getting a disturbing
> > result.  Yesterday, for instance, my realtime strategy
> > and corresponding indicator was flat the whole day.
> > Today, after rebooting, I loaded the same exact system
> > using the same data (except now historical).  It
> > claims to have given me a nice sell signal midday,
> > just before the nasdaq tanked.  In fact, the realtime
> > signal/indicator are totally different!
> >
> > Have others noticed anything like this?  At first I
> > thought it might be due to precision errors.  But I am
> > not sure that make sense since I thought the chart is
> > recalculated each bar from scratch.
> >
>
> <snip>
>
>