Announcement

**ACM** · 06-30-2008, 06:09 AM

Re: Closing Price Discrepancies

Dave
That is because Daily (and longer) intervals include the Settlement which is transmitted after the regular trading hours close
Alex

Originally posted by Dave Skram
Can anyone share some insight on why there are differences between the end-of-day closing values for Daily and longer (Weekly, Monthly, etc.) intervals versus intra-day (60, 30, 15, etc.) intervals?

I’ve attached a spreadsheet file from this morning showing the values for last Friday’s close for various intervals for QQQQ and QID. The difference for QQQQ is only $0.01 ($45.65 versus $45.64) but the difference for QID is $0.21 ($43.86 versus $44.07).

This server connection information is for QLink 1.0b2:

Proto Foreign Address State
TCP 216.23.234.119:2189 ESTABLISHED
TCP 216.23.232.177:2192 ESTABLISHED
TCP 216.23.232.236:2194 ESTABLISHED
TCP 216.23.224.98:2193 ESTABLISHED
TCP 216.23.226.191:23100 ESTABLISHED
TCP 216.23.226.191:23100 ESTABLISHED

-Dave

**Dave Skram** · 06-30-2008, 06:34 AM

Thanks Alex. I can understand the $0.01, but a difference of $0.21 seems a bit excessive.

How would you recommend handling calculations that integrate multiple intervals that need to span that settlement threshold?

This particular issue is impacting an automated system that needs to reconcile the closing values from 60-minute, Daily and Weekly intervals to initiate actions.

Differences of a cent or two don’t impact the outcomes in a material way. But amounts of 0.5% are a problem.

Is the answer to build the Weekly and Daily OHLC values from the 60-minute data? That seems to me to be the only way to have the data from all three intervals be consistent.

-Dave

**Larry Marchman** · 06-30-2008, 06:36 AM

Hey Dave,

What Alex says is so. To be redundant here, the intraday quits at the last intraday RTH bar (in QLink 1) and the final price can dribble in a little after close which will show in Daily(and longer).

There's an additional specific problem(bug) though with QID that I just happened to notice when looking at it's T&S. It's currently carrying the wrong intraday RTH close of 16:00. That symbol should be RTH close of 16:15. This is contributing to a wider difference in price than it should.

I'll get the MktHours guy to fix it. Most likely it will be in the July 4 holiday update due in a couple of days.

LAM

**ACM** · 06-30-2008, 08:48 PM

Dave

Is the answer to build the Weekly and Daily OHLC values from the 60-minute data? That seems to me to be the only way to have the data from all three intervals be consistent.

You could also try using a 390 minute intraday interval [for NYSE or Nasdaq equities] as a proxy of the daily and then build the weekly from that.
Alex

Originally posted by Dave Skram
Thanks Alex. I can understand the $0.01, but a difference of $0.21 seems a bit excessive.

How would you recommend handling calculations that integrate multiple intervals that need to span that settlement threshold?

This particular issue is impacting an automated system that needs to reconcile the closing values from 60-minute, Daily and Weekly intervals to initiate actions.

Differences of a cent or two don’t impact the outcomes in a material way. But amounts of 0.5% are a problem.

Is the answer to build the Weekly and Daily OHLC values from the 60-minute data? That seems to me to be the only way to have the data from all three intervals be consistent.

-Dave

Announcement

Closing Price Discrepancies

Closing Price Discrepancies

Comment

Comment

Comment

Comment