[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

GLOBEX T&S vs eSignal ticks



PureBytes Links

Trading Reference Links

Hi, I parse out CME data myself and am familiar with your issue. My vendor 
of comparison is Realtick vs. eSignal. Just a note of caution regarding 
intraday data. ALL VENDORS DIFFER.

Miles Dunbar presented a detailed analysis at a Futures symposium that 
compared intraday data from different sources. It turns out that even the 
same vendor across cable, satellite, and network varied on bars, including 
highs and lows!!! The differences across vendors is even greater. Larry 
Williams likes to point out the same thing, and I've certainly witnessed 
real-time differences first-hand.

This has an VERY important impact on intraday back testing. You will not 
get the same signals on CME data as eSignal. The CME data is valuable to 
get a general idea of system performance, but backtest on eSignal data to 
get something closer how it would really trade for you.

What I've done is strip off and save Realtick data for the markets I trade, 
then backtest on that data. I do use CME data exactly as I suggested.

Cheers,

Kevin

> From: 9m9@xxxxxxx
> To: omega-list@xxxxxxxxxx
> Subject: GLOBEX T&S vs eSignal ticks
> Message-ID: <8425904718.20031115202702@xxxxxxx>
> Content-type: text/plain; charset=us-ascii
> Content-transfer-encoding: 7bit
> Hello omega-list,
> I try to figure out how data that i use (eSignal) correspond with
> GLOBEX data.
>
> I parse GLOBEX Time and Sales datafile 
(ftp://ftp.cme.com/pub/gtime/gx031114.zip)
> and got 54653 T&S records. Then i sum up volume from all records and
> got total 630640 contracts (i compare just day session 9:30-16:15 ET).
> I opened 1 tick chart in TS and inserted Volume indicator and export
> data to file. I parse file and got 109185 records then sum up all
> volume from all records and got 622679 contracts (day session).
> Now question is why the difference in number of records between GLOBEX
> T&S file and tick file from TS (with eSignal) so big (54653 vs 109185)?
> Volume about the same 630640 and 622679.
>
> Or maybe i use a wrong method to compare data?
>
> TIA