|
PureBytes Links
Trading Reference Links
|
But we're not talking about the ever-present bad ticks that occur. We're
talking about a design flaw that allows individual bad ticks to destroy
an entire block (day?) of data. It's a different order of magnitude.
A more appropriate design would accept (or reject!) a bad tick without
wreaking havoc on all subsequent data. When your entire data stream
becomes corrupted because of one bad tick, it's tough to be a good
trader.
BTW, this is intended to be a valid criticism in search of a better
design solution. Flaming is inappropriate.
Allan
___________________________________________________________
Mark Seleznov wrote:
>
> Although I am a Omega Solution Provider, I really need to come to the
> defense of Omega on all this data bashing.
>
> There are errors in everyone's data. IT IS ALL BAD! I am the Managing
> Partner of a stock brokerage firm. We have S&P data feeds, BMI data feeds,
> North American Quotation data feeds, and use data feeds from other providers
> for clients too. The exchanges distribute the data to these providers. The
> data providers only broadcast what the exchanges send them.
>
> The data is bad and I have seen some crazy stuff over the years. For
> example, a Market marker enters 7 1/2, instead of 77 1/2 and a trade takes
> place.
>
> I was a market maker on the Phila Options Exchange, the clerks would mistype
> prices into their machines. This happens all the time. It happens on all
> exchanges.
>
> The most critical issue which traders do not realize is that under exchange
> rules, a firm has 90 seconds to report trades. Many trades that take place
> on a principal basis are delayed and are out of order. Big blocks get held
> or stopped and reported many, many seconds later. You see this all the time
> on Level 2 quote screen when the market is 1/8 X 3/16, and a 5/16 block of
> 25,000 shares prints. The trade actually took place a minute or so before.
> That is tape reading and you can not blame Omeag or data vendors.
>
> Exchanges, ECN's, specialists, market makers all have data reporting
> problems that contribute to unreliable data.
>
> I always laugh at the mental exercises that some of you play on tick data.
> It is a joke.
>
> Omega is a great product. Every product has limitations and the data
> bashing is not an excuse for poor trading techniques.
>
> Mark
>
> Mark A. Seleznov
> Managing Partner
> Trend Trader, LLC
> Member NASD, SIPC
> 15030 N. Hayden Road
> Suite 120
> Scottsdale, AZ 85260
> 602-948-1146
> 602-948-1195 fax
> mark@xxxxxxxxxxxxxxx
> http://www.trendtrader.com
>
> -----Original Message-----
> From: greene@xxxxxxxxxxxxxxx <greene@xxxxxxxxxxxxxxx>
> To: Allan L. Kaminsky <allank@xxxxxxxxxxxxxxxx>
> Cc: IUhrik <IUhrik@xxxxxxx>; omega-list@xxxxxxxxxx <omega-list@xxxxxxxxxx>;
> officeofthepresident@xxxxxxxxxxxxxxxxx
> <officeofthepresident@xxxxxxxxxxxxxxxxx>
> Date: Tuesday, January 13, 1998 5:26 PM
> Subject: Re: Letter to B. Cruz on bad tick editing
>
> >The normal rule of thumb is a product is defective if it's not reasonably
> >suitable for its intended use. I'm not a day trader - but any program that
> >would wipe out large chunks of my database without asking me first sounds
> >pretty defective to me <g>. OTOH - since you know about the problem - it
> >probably wouldn't be a great lawsuit if you tried to claim a trading loss
> >(depending on what state you live in - I assume you live in the US -
> there'd
> >be issues of comparative negligence). On the third hand - I think you'd
> have
> >a much better case if all you were seeking was a refund for the price you
> paid
> >for the program.
> >
> >BTW - I have no particular ax to grind when it comes to Omega. I just
> resent
> >spending a lot of money for any software that doesn't work properly (and
> most
> >software manufacturers don't seem to have any compunctions about bringing
> >products which are basically beta versions to market). Robyn
> >
> >Allan L. Kaminsky wrote:
> >
> >> Well written letter. Hope you get a rational response.
> >>
> >> Kind of makes you wonder about product liability. A product sold for the
> >> express purpose of trading is unable to protect the user from common
> >> occurrences (not unforeseen acts of nature). Is there product liability
> >> if the user then suffers losses as a result of the poor engineering of
> >> the product? In many other areas, the manufacturer would be at great
> >> risk (aircraft manufacturers, etc.).>>
> >
> >
> >
> >
|