|
PureBytes Links
Trading Reference Links
|
At 02:58 PM 11/29/2004, Doug Tucker wrote:
>Does anyone understand why the s&p e-mini data is so different on
>tradestation 8 as compared to DTN or most other feeds. There are about twice
>as many ticks with TS8. I compared DTN with TS8 for about a month and
>counted the daily ticks using each. TS8 has between 1.9 and 2.1 times as
>many ticks.
At 05:19 PM 11/29/2004, Jimmy wrote:
>How could they work around it when you refresh or have to refresh a
>chart and the ticks get cut in half. All new bars. Your 100 tick
>chart becomes a 50 tick chart. Not good.
This is not true. Please get your fact straight before jumping to conclusion.
This has come up before.
The conclusion was that CME does not report all the ticks and TradeStation
and ESignal do.
BobR and I ran tests on this in July.
Collecting the data real time I received 64,464 ticks of ES from 9:30AM
through 4:15PM Eastern time.
Refreshing the data, I then had 64,497 ticks, a difference of +33 ticks.
So I got 33 more ticks vs. real time when I refreshed the data after the
close of the day, not half the ticks as you said. This is very good
consistency, in my opinion.
Details below.
Bob Fulks
-------------------
I ran a test for Friday, 7/2/04.
Collecting the data real time I received 64,464 ticks of ES from 9:30AM
through 4:15PM Eastern time.
Refreshing the data, I then had 64,497 ticks, a difference of 33 ticks.
So refreshing the data did not eliminate any ticks but got 33 more.
So it appears that few ticks were lost in transmission. This is with a Cable
ISP. Certainly very good by any standard.
BobR checked CME and ESignal for the same day:
BobR wrote:
With esignal on 7/2 I received 62167. The eod download
produced 64342. Not a quality situation for realtime data. My DSL is on
two wire and at the limits of the DSL range. There are 13 hops between me
and the esignal server with a 39ms roundtrip time. I have the procedures
for getting a CME count and will try that this weekend.
BobR regrding CME count:
The last line shows the row count of 33651. Strangely, both esignal and
TS7(8) both give readings twice that of the CME. This has been true ever
since I switched from dtn sat to esignal dsl in Dec 2003. No one has
offered an explanation of the 2X reading. What I don't know is if esignal
and TS are correct and the CME count is wrong......? We assume their reading
is gospel, but who knows, maybe they are wrong, seems impossible.
Bob Fulks:
I got a call from someone at TradeStation. The person told me that the CME
leaves out some ticks for some reason so the CME reported values are much
less than the actual number of ticks reported by TradeStation and other
vendors.
He was expecting a message from CME explaining this and said he would send
it to me when he received it.
BobR Replied:
Sounds right. I know another trader that made a tick by tick comparison and
figured out exactly what the CME is doing with their data. He did it in
excel and sent me a picture. I was busy at the time and didn't try to
understand what he was saying. That was a few weeks ago. I'll retrieve the
email and try to understand it. I emailed both the cme and esignal and was
surprised neither answered correctly. What bothers me is dtn has pretty
much the same number of ticks that cme has and with esignal having twice,
this has to have charting/trading implications when using tick charts. It
is more complicated than just posting both sides of the trade. I would
still like to hear the correct answer from the cme and esignal. Their rep
on the omegalist hasn't offered a single explanation and my email to their
company resulted in an faq referral with no information. Truth is I haven't
pursued either of them enough times to dig out the answer.
|