|
PureBytes Links
Trading Reference Links
|
1/2 second resolution simply does not exist in retail, regardless of
the data feeds and software products available to us. A T1/T3 line
directly to the CME is better; buying yourself a seat at the exchange
is best.
Those of you who are working on systems that scalp split-second
teenies several times a day may worry about other problems, like your back testing
protocols that distort the reality of slippage and actual fills.
Wednesday, June 18, 2003, 6:11:28 AM, you wrote:
AP> Mr Greetings,
AP> Are people not kidding themselves if they think they can scalp the S&P
AP> via quote machines?? I'm sure successful machine based scalpers..and by
AP> scalpers I mean true scalpers, would be as rare as hens teeth.
AP> Are you also suggesting that when you know your system is 1/2 a second
AP> slow, that it causes you emotional angst? Even though in practical
AP> terms it might make little difference to your trading? That doesn't
AP> make much sense to me. Thats like me making a $1000 on a trade and
AP> getting annoyed that I traded through a broker that charged me $7 r/t
AP> instead of $6!!
AP> You failed to say what the net difference the 50% different fills made
AP> to your system and over what number of trades? I'm thinking the alert
AP> sound on TS would last longer than a 1/2 second!! LOL.
AP> Regards,
AP> Adrian
>> -----Original Message-----
>> From: trader@xxxxxxxxxxxx [mailto:trader@xxxxxxxxxxxx]
>> Sent: Wednesday, 18 June 2003 7:31 PM
>> To: omega-list
>> Subject: Re: New TS CME DataFeeds, a Quick Review
>>
>>
>> ----- Original Message -----
>> From: "Adrian Pitt" <apitt@xxxxxxxxxxxxx>
>>
>> > People complain about TS6/7's slowness in ES quotes, but
>> does 1/4-1/2
>> > a second really matter when all is said and done? If
>> someone blinks
>> > at a different time to someone else, that covers the
>> difference!! Lol
>> > In an automated system your order 'may' get in line quicker, but
>> > unless your dealing in size does it really matter? I get
>> the feeling
>> > people are talking semantics when it all boils down to real
>> trading.
>> > Especially as most don't do automatic order routing. Am I missing
>> > something here?
>>
>> Yes, you are missing that not all people trade alike.
>>
>> Where one trades EOD or Weekly systems, half a second is not
>> important. But some people scalp and have to make just a few
>> points of profit per trade.
>>
>> Now maybe in reality the 1/2 a second would not make a
>> difference and in the end, after a years trading, the
>> differences add up to nothing (average out is the right word
>> i believe)
>>
>> For me, IF i would trade a system like that (wich i do not by
>> the way) there would be a much bigger problem being that you
>> are not right on to it. External factors are influencing your
>> decissions and that is something you don't want. There are
>> many people out there and if lagging data is setting you a
>> step behind, that gives a very shitty feeling. Emotions about
>> things like that would be very high with me. You don't want
>> external factors messing around with you, trading can be hard
>> enough as it is allready.
>>
>> As said, i don't trade like that, but even with longer time
>> frame systems (i trade 15 minute systems) i don't want to be
>> taken by suprise. Half a second can make a difference.
>> Actualy, i am convinced it does. I have two systems running
>> both TS but with two different datafeeds. pc's are set to
>> synchronize with an atomic clock each night but one of the
>> datafeeds is lagging, just a very small bit. The trades are
>> in 50% of the cases executed at a different price. Most of
>> the time only a 1 point difference (future trades at 2500 so
>> 1 point is not
>> much)
>> but there you see that lagging can make a difference.
>>
>> Greetings
>>
>>
|