| Author |
Topic: Volume Discrepancy (3 messages, Page 1 of 1) |
||||
|---|---|---|---|---|---|
| Moderators: VPfau | |||||
|
NVerner2 Posts: 12 Joined: Apr 17, 2012 |
Hello,
I received some NQM2 sample data from CQG, one of the most reliable market data providers in the world, and their volume data were quite different from the data I downloaded using the OEC Demo API. I acquired the OEC data by using a RequestBars command. The data from both CQG and OEC are one-minute bars. The average volumes per minute for the NQM2 for CQG and OEC were 18.6 and 239.1, respectively, from May 13 to May 15. Do you know why there is such a large difference? The other data from CQG seem quite close to what I have on OEC. I only saw 15 total errors in the Open, High, Low, or Close for roughly 2000 minutes. The errors were quite small, no greater in magnitude than 0.25. However, I don't understand why the volumes would be so different for the same contract. I would appreciate any insight you have on this issue. Thanks. Nathan verner & Fred isabelle |
||||
|
VictorV Posts: 746 Joined: May 08, 2007 
|
Hello,
could you show a source code that you are using to compare bars? What time of day do you use to compare bars? Did you compare a sum of one-minute volumes for day between CQG, OEC and exchange? Victor Vins Lead Software Developer |
||||
|
NVerner2 Posts: 12 Joined: Apr 17, 2012 |
There really isn't any source code to compare bars. I just copied and pasted all the data into an Excel spreadsheet and subtracted the OEC data from the CQG data. I had about 2200 data points to use for a comparison, from the Open on Sunday to the Close on Tuesday, so all times of day were seen in the data. The OEC volume data are off by a factor of 10 from the CQG data, no matter the time of day.
I don't have exchange data to compare with the other two. I'm not sure how I would obtain that. I was able to obtain some data from another provider named TickDataMarket. Their volume data were quite close to yours, but not exactly the same. There was still an error rate of approximately 26% in the volume data between OEC and TDM. In contrast, the error rate was approximately 97% in the volume data between OEC and CQG. I asked CQG about this problem, but they have not responded. I'm not sure what the issue is, as CQG has quite a good reputation for quality, but your data agree with TDM quite well. Nathan verner & Fred isabelle |
||||