|
PureBytes Links
Trading Reference Links
|
I have to apologize for blaming the low tick counts upon DBC. The
source of the "problem", if there is one, turns out to be the CME
itself.
Early this year I extensively researched the content of the time-&-sales
and volume-by-tick files available on the CME's FTP site, including
detailed conversations with several CME technical people. At that time,
the difference between these two files, after accounting for duplicate
prices, bid/ask entries, cancellations, insertions, etc., was very
small. Because I was interested in playing with intraday volume, I
decided to collect and retain the vol-by-tick files only, filtered to
eliminate dupes, etc.
My recent complaints directed at DBC about low tick counts was based
upon comparisons between collected data (mine & others') and the
filtered vol-by-tick files. This afternoon, however, after all the
e-mail I received suggesting that I might be wrong, I compared my
collected data with CQG data from my broker, and with both the CME T&S
and V-by-T files.
I found that my collected data for 9/14, after a correction estimate to
account for the 7- or 8-minute outage, was 2574 ticks; the CME figures
are: 2711 for their T&S file, and probably about 4800 for vol-by-tick
(that file is not yet available for 9/14).
The question THIS now raises is: Why is the count of ticks coming from
the CME now so low??? The vol-by-tick counts are certainly much more in
line with counts we were receiving six months or more ago!
Regards,
Carroll Slemaker
|