|
PureBytes Links
Trading Reference Links
|
I once compared futuresource satellite to esignal to ts. I found that not
only was timing different, but the values were different. Futuresource
seemed to be getting its data from a different source than the others.
Different sampling rate or something. Same for upvol dvol data. A system
that I have using uvol and dvol is hugely profitable on futuresource but not
on the others.
Chris
----- Original Message -----
From: "carrslem" <carrslem@xxxxxxx>
To: "The Funkhousers" <funkhouser@xxxxxxxx>
Cc: "Compra dun" <compradun@xxxxxxxxxxx>; <david_colin@xxxxxxx>;
<omega-list@xxxxxxxxxx>
Sent: Monday, February 24, 2003 6:31 PM
Subject: Re: Oddball oddities
> This would produce a more-or-less systematic difference. But I have
noticed
> two different anomalies: (1) a small very short-term (a few minutes)
random
> fluctuation over the course of a day, and
> (2) a situation where the two feeds were nearly identical on most days,
but
> with occasional days when the sum of advances and declines was nearly
> identical but the difference (advances minus declines) differed between
the
> two feeds by between 50 and 100. Furthermore, this interfeed difference
> remained virtually constant throughout the entire day, but the next day
all
> was well again.
>
> Case #1 is to be expected. Case #2 is very strange - it is as if one of
the
> feeds has transposed a set of particular issues from the adv category to
the
> decl category (or vice versa) and retains that transposition all day.
>
> Carroll Slemaker
>
>
>
> > Advance Decline statistics differ among the various providers due to the
> > definitions that vary among them.
> >
> > Many publish the gross advance declines (including all of the securities
> traded
> > on the exchange, regardless of type) and others remove the advance
decline
> > numbers the issues which are not strictly pure individual equities,
i.e..
> > preferreds, etc., on the assumption that changes due to interest rates
> and/or
> > other factors are not indicative of equity changes.
> >
> > Richard Funkhouser
>
|