|
PureBytes Links
Trading Reference Links
|
The following is some additional clarifying information about the report
posted on this subject.
The statistics are all based upon data for the September S&P futures
contract. This symbol was chosen because it is the only one I know of for
which a virtually absolutely accurate file of data is available for free.
Such a file is necessary to serve as a standard against which to compare
data from a feed.
For the purposes of this study, a trading day is divided into a series of
half-hour periods and statistics are collected independently for each
half-hour slot.
For tick counts, I consider the most significant statistic to be the
"ratio".
A "gap" is an instance where one or more consecutive ticks reported by the
CME are missing from the feed. For each half-hour slot, data are given for
only a single gap (if any occurred), the worst one (longest time). For that
worst gap the following are given: the length of time during which no ticks
were received from the feed, and the number of ticks missed in that gap.
For example, in the table of data for Signal/cable for 6/30 at 11:30, length
is shown as 50 and ticks is shown as 8. This means that during the half
hour ending at 11:30 PT the worst gap lasted 50 seconds and lost 8
consecutive ticks. Note that altogether 83 ticks were lost during that half
hour so there were other gaps - the statistics describe only the worst one.
If anyone can contribute intraday ASCII data for PCQuote, or for any other
feed not mentioned in the report, I'd be happy to run it through my analysis
program and report the results. Beause of the dubious nature of the ESignal
results so far, I would also appreciate receiving ESignal data from other
subscribers, especially subscribers using a reliable broadband connection to
the Internet.
Regards,
Carroll Slemaker
|