re: Order Cancel Request
Anonymous
24 Feb 1998 4:06PM ETVitaliy Kantor wrote:
> FIX 4.1 does NOT ignore partial cancels. They can be expressed in Order Cancel/Replace message without any problem whatsoever.
>
> Therefore NO functionality is lost.
>
That is technically true. However, to interface a FIX
system to a CMS system would therefore require the system
to map the Order Cancel/Replace message (when appropriate)
into a partial cancel message. This is possible but
messy. This approach obscures the actual process from the
traders, and can therefore lead to much confusion in the
event of difficulties. On a business level, there are real
semantic differences between partial cancels and order
cancel/replacements, and the best implementation would
transparently reflect this.
Please remember: the underlying issue here (and the reason
for the development of the partial cancel to begin with)
is that a partially cancelled order retains its time priority on the order book (most books operate on the
basis of a price/time priority system). Any other type
of cancel/replace results in the time being reset to the
current time, with the concomitant loss of position on
the book (it goes to the back of the queue). This is a real
business issue, and a good technology solution would not
obscure this!
> The discussion is only about implementation of partial cancels and although Andy's arguments are quite reasonable there are counter-arguments at least of the same level of reasonability.
If you have such reasonable counter-arguments, please
present them.
>
> Just rest assured as far as business is concerned you can and may convey partial cancels in NYSE-way through FIX 4.1
Yes and no. Such a statement trivializes the issue, in
my opinion. Remember, I'm not asking you to use this
feature. I'm just asking you to admit it to the protocol.
In view of all the other stuff in the spec, it is
mind-boggling to me that this feature should encounter
such resistance.
Andrew Schorr
Senior Vice President
Daiwa Securities America Inc.
