Data Integrity - Splits, gaps, spikes, whatnot.

[bobscott]

Bill Bledsoe maintains a website devoted to QCharts and QCharts tools. Bill is kindly hosting the QFeed Application Interface (QFAI) which was written by QCharts users and includes use of a fairly detailed "splits" file CorpActs.UNF which I maintain. When running QFAI, the CorpAct.UNF file will "trump" the normal QCharts CorpActs.INF file. Details on QCharts useage and QCharts tools may be obtained from here .

~Bob

[ckelly]

Mattman,

Additionally, you can go to: www.investmenthouse.com and see their free Splits Calendar.

Carol

[bpschultz]

Bad Data for GC07G

There is a major problem with the symbol GC07G, which is the front month gold futures contract. I submitted a request to address the problem via a phone call to tech support about a week ago. Since nothing has been done, I thought I'd raise the issue here.

The problem is evident using a 60 minute, all sessions chart. The regular trading hours show correct prices, but the electronic trading hours prices are obviously quite wrong.

I would appreciate an immediate fix to the problem, since my previous request which was made prior to this contract becoming the spot contract was not addressed in a timely manner. Thank you.

[Larry Marchman]

Re: Bad Data for GC07G

Hey Brian,

I see the issue you report. Of course that's beyond my purview.

However, I would like to comment regarding "front month". I see that NYMEX.com, on the home page, seems to indicate that GC07G (Feb. 2007) is the front month according to the right-side mini chart display when one mouses over the various symbols.

However, in the real world, traders consider December 2006 as the front month. A comparison of the actual trading volumes will show this. The de facto front month is the most actively traded contract, from a trading and liquidity POV. Also, eSignal has Dec 2006 as front month with their continuous contract symbol (GC #F) and their front month symbol (GC 1!).

Just a friendly fwiw here.

LAM

Originally posted by bpschultz
There is a major problem with the symbol GC07G, which is the front month gold futures contract.

[bpschultz]

Gold Front Month

It is not my intent to be argumentative here. However, as someone who trades the precious metals for a living, I would note that as of last Wednesday, the majority of the open interest in the Comex gold futures was in the February 2007 contract. Also, some brokers will no longer allow their clients to trade the December contract due to the proximity of first notice date. Most traders who have a futures position in gold have already rolled forward to the February contract and are currently trading this contract.

[JayF]

Thanks Brian for the report. We have been working on some general fixes for NYMEX, COMEX and NYBOT that have been tested and are in the process of being rolled out to servers. These fixes involve some changes to the symbology of future option symbols, which will be laid out in a KB article when the release is completed.

[Larry Marchman]

Re: Gold Front Month

Hey Brian,

Originally posted by bpschultz
Most traders who have a futures position in gold have already rolled forward to the February contract and are currently trading this contract.

It's also my intent to not be argumentative. I only seek clarity. I don't trade GC, but I do trade Futures (not Options on Futures). Namely ES/NQ/MR/YM. As a daytrader, liquidity is the thing, and the front month has the liquidity. On a roll day, the volume switches to the new front month. For the CME emini's this day is well defined, and the volume switch is most apparent. I also see this volume switch in GC in looking at past contracts a day or two before the end of the month prior to the month of expiry. It looks like GC06Z may roll to GC07F in a couple of days, but not as yet, as best as I can tell. Again, by my needs and my definition, roll means volume.

The screenshot here is RTH with interval of 310 (8:20-13:30) to show a single bar reflecting a given day's volume. It's just an easy way to view roll.

Hope you can appreciate why I mentioned GC06Z as of today being the fronth month, since the vast majority of trading is on that contract.

LAM

[bpschultz]

Gold Front Month

Jay,

What is the status of the repairs for the GC07G contract? The all sessions chart is a horror to look at and is completely unusable. I think Larry would concur that it is now the front month. How am I supposed to trade under these circumstances? Thank you.

[Larry Marchman]

Re: Gold Front Month

Yep, it no doubt rolled today. And also yep, it's a horror All Sessions.

Last Friday the 24th intraday RTH is also a mess. Since you're a knowledgable MktHours hacker, and I'm sure may have already done it, coding the 24th as "H" would at least clean up the intraday RTH charts.
Yeah, I know, little consolation.

LAM

Originally posted by bpschultz
Jay,

What is the status of the repairs for the GC07G contract? The all sessions chart is a horror to look at and is completely unusable. I think Larry would concur that it is now the front month. How am I supposed to trade under these circumstances? Thank you.

[bpschultz]

Thanks, Larry.

[bpschultz]

Updates to Mkthours.ini File

Jay,

I know how you guys love the mkthours.ini file, so I wanted to point out that changes are needed to the Comex precious metals symbols GC, SI, and probably others, as a result of the beginning of round-the-clock electronic trading on Globex.

From the Comex website at http://www.nymex.com/GC_spec.aspx:

Effective December 3
Electronic trading will be conducted from 6:00 PM until 5:15 PM via the CME Globex® trading platform, Sunday through Friday. There is a 45-minute break each day between 5:15PM (current trade date) and 6:00 PM (next trade date).

Note the change in the end of the current trade day from 14:00 to 17:15 on Mondays through Thursdays. Also note that trading no longer halts from 8:00 to 8:20 and 13:30 to 14:00 Mondays through Fridays. These updates need to be reflected in the mkthours.ini file.

Thanks for making these fixes.

[bpschultz]

Oops!

Jay,

Sorry for the earlier post about the mkthours.ini file. I forgot that I've been using a markethours.unf file and now see that eSignal has already made the changes in the mkthours.ini file. Thanks.

[Larry Marchman]

Symbol list on servers incorrect for some syms

The following issue (see paste below) was reported on the QCharts Yahoo!Groups forum.
See LINK for initial problem reports and my reply.

Could someone please get this info to the proper dept.
Thanks.

LAM

************** paste of my reply to YahooGroups *********

Looks like the symbol list on the servers has gone bad for some syms. The list on the server contains both Exchange and Symbol. For IWM, the servers have NYSE:IWM. It should be AMEX:IWM. One doesn't see the issue in QCharts if one is clicking on a quotesheet that had been created before the recent change to the symbol list on the servers. The symbol field in a quotesheet has both Exch:Sym, and a quotesheet that existed before the current change will have AMEX:IWM in it, which will yield the correct data in QCharts. I have no clue if Exchange:Symbol can be specified in QLink, or what one could do as a workaround for any given 3rd party app.

Here's a few symbols I found that have mis-assigned Exchanges.
This is just from a quotesheet I have containing a partial set of ETF's, and is hardly all inclusive.
btw. IWM, etc are not Indexes, but tradable ETFs/iShares

All these are assigned Exchange NYSE, and should be assigned AMEX
IWM
IWN
IWO
EWG
EWW
IWB
IWD
IWF

I've reported this issue on the eSignal forum.

LAM
**************** end paste **********************

[bpschultz]

DX00Y is working on some servers and not others.

GC07G has been partially restored on some servers and not others.

As fate would have it, I haven't found a server on which both DX00Y and GC07G are both working. There probably are one or two, but why should I have to take the time to go through each and every server? What happens when that server is busy and won't let me on it?

The history on GC07G is still too dirty for the symbol to be usable. I use long-term moving averages on my all session, intra-day charts which are still rendered useless by the bad data which goes back hundreds of bars using a 120 minute, all sessions chart. The clean data only goes back a week or so, which is hardly useful even without the moving averages.

Are there any plans to address these issues? Thank you.

[Larry Marchman]

GC07G data

Hey Brian,

Regarding the 120 ETH you mentioned, a fwiw regarding a workaround if you're interested.

That chart will perhaps be useable if you do a continuous contract. The only drawback is the inherent price gap at Roll date. Find below a copy of my continuous for GC that goes back to GC05J. This is suitable for a select/copy/paste into your CompSyms.ZNF. The pasted result in your ZNF should be two lines when viewing in Notepad with word wrap Off. It looks w***** here due to browser, but a select/copy/paste from here into CompSyms.ZNF via Notepad will be two non-w***** lines. The // is just reminder comments I have for me. The symbol to use in quotesheet or interval box would be GC.Q .

//GC months... Z,Q,M,J,G RTH Hours 8:20-13:30 = 310min. Use Interval 310 to see actual volume shift. "V" is active, but skipped here. Front month goes from Q to Z and clearly so.
GC.Q ; COMEX:GC.Q ; Gold "ZNF" ; COMEX:GC07G ; DT(#COMEX:GC07G|11/29/2006|-) DT(#COMEX:GC06Z|7/28/2006|11/29/2006) DT(#COMEX:GC06Q|5/29/2006|7/28/2006) DT(#COMEX:GC06M|3/29/2006|5/29/2006) DT(#COMEX:GC06J|1/30/2006|3/29/2006) DT(#COMEX:GC06G|11/30/2005|1/30/06) DT(#COMEX:GC05Z|7/28/2005|11/30/2005) DT(#COMEX:GC05Q|5/25/2005|7/28/2005) DT(#COMEX:GC05M|3/22/2005|5/25/2005) DT(#COMEX:GC05J|1/28/2005|3/22/2005)


LAM