|
PureBytes Links
Trading Reference Links
|
Bill Vedder wrote:
> Is it the 'tick count' that's important? Or is it the time the tick
> arrives that's important?
[This was in response to my earlier remarks: "I am using Signal,
delivered by cable. Some time back I compared a full day's tick count
for the front S&P contract; I compared my count with that of a friend
using BMI and with a count from the official list of ticks from the CME
itself. My recollection is that my count was very close to the BMI
count and both were somewhat lower than the CME's count."]
Actually, "somewhat lower" is a considerable understatement. I just did
a comparison between my count and that of the CME for 4/29 and 4/30.
Here are the results:
4/29 - My count = 4336, CME = 5664
4/30 - My count = 4293, CME = 5470
(These are for the day session only - GLOBEX is not included.) This
represents an average tick loss of 22.5%, a VERY significant loss.
As for arrival times, past comparisons with the CME file indicated a
delay of no more than a small fraction of a minute most of the time.
Carroll Slemaker
|