I was checking some exhaustion volume stuff on rangebars and noticed some weird results. I noticed that I was getting hundreds of millions of contracts per 5 point rangebar on the ES today (20090218). I promptly realized that I have finally gone totally nuts. I have checked using reserved words; upticks, downticks, ticks and they are all way out of whack.
Can someone pull up a 5 point range bar chart on the ES and check this? (see attached) I have normal volume on tick bars, volume bars, minute bars, etc... just not on range bars.
Thanks!
(yes, that's the way my tool bars always come up when the program starts. I just didn't re-arrange them for this screen shot)
Range Bar Volume
-
- Posts: 407
- Joined: 01 Jan 2008
- Been thanked: 3 times
- Bruce DeVault
- Posts: 438
- Joined: 19 Jan 2010
- Location: Washington DC
- Been thanked: 2 times
- Contact:
As a philosophical matter, if everything is working properly in the platform (and as importantly, if eSignal's tick data matches perfectly with their minute data, which isn't always the case), we would expect the sum of the volumes for the day to be the same regardless of what kind of bars we use.
So, let's compare this between 5 point bars of ES and 1 minute bars.
In the attached screenshot, you can see a volume spike similar to the one you showed at a particular 5 point bar. Because price didn't move very far up or down and only sideways for a period of time (during which there was normal volume), that volume accumulated until finally when price did exceed the 5 points, the accumulation was more than normal.
In the attached screenshot, blue represents the 5 point bars and their volume and accumulated volume, while orange represents 1 minute bars and their volume and accumulated volume. As you can see, while they don't match exactly (most likely because the eSignal data is different between ticks and minutes), they do match closely, including at the time of the spike in question.
The example DayVol indicator shown here has only the following line of code, and is intended only as a quick example:
In some cases as you scroll around with a validation tool like this one, you may notice that volume on one side lags the other, which could be indicative of a data issue, or possibly of a platform issue, but I would always suspect the former before the latter until proven otherwise, because data issues of this sort are widely known and expected. That doesn't mean there's not a platform issue, only that you should always first rule out differences in the data so you can report more specifically the nature of any issue in the platform, because data differences between tick and minute happen all the time and are normal.
You may find it instructive to take this same workspace and change the 1 minute data series to 1 tick, then pull your bar spacing very tight, and see how the situation stacks up at particular times, as this method will isolate exactly what the platform is doing with volumes in the range bars.
So, let's compare this between 5 point bars of ES and 1 minute bars.
In the attached screenshot, you can see a volume spike similar to the one you showed at a particular 5 point bar. Because price didn't move very far up or down and only sideways for a period of time (during which there was normal volume), that volume accumulated until finally when price did exceed the 5 points, the accumulation was more than normal.
In the attached screenshot, blue represents the 5 point bars and their volume and accumulated volume, while orange represents 1 minute bars and their volume and accumulated volume. As you can see, while they don't match exactly (most likely because the eSignal data is different between ticks and minutes), they do match closely, including at the time of the spike in question.
The example DayVol indicator shown here has only the following line of code, and is intended only as a quick example:
Code: Select all
variables: dv(0); if date<>date[1] then dv=volume else dv=dv[1] + volume; plot1(dv, "DayVol");
You may find it instructive to take this same workspace and change the 1 minute data series to 1 tick, then pull your bar spacing very tight, and see how the situation stacks up at particular times, as this method will isolate exactly what the platform is doing with volumes in the range bars.
- Attachments
-
- Volume Comparison Screenshot.png
- Volume Comparison Screenshot
- (188.66 KiB) Downloaded 1845 times
-
- Volume Comparison.wsp
- Example workspace
- (37 KiB) Downloaded 534 times
Does eSignal have filtered or unfiltered ticks or snapshots in their data feed?
I once looked at OEC forum for developers and it turned out that OEC filters ticks / throttles them and from this throttling people get unrealistic volume in different platforms depending on what type of bars are used for charting in third party platforms.
I once looked at OEC forum for developers and it turned out that OEC filters ticks / throttles them and from this throttling people get unrealistic volume in different platforms depending on what type of bars are used for charting in third party platforms.
- Bruce DeVault
- Posts: 438
- Joined: 19 Jan 2010
- Location: Washington DC
- Been thanked: 2 times
- Contact:
- Bruce DeVault
- Posts: 438
- Joined: 19 Jan 2010
- Location: Washington DC
- Been thanked: 2 times
- Contact:
- RobotMan
- Posts: 375
- Joined: 12 Jul 2006
- Location: Los Altos, California, USA
- Has thanked: 31 times
- Been thanked: 13 times
- Contact:
Thanks Tekram!
Thanks for the suggestions. I use other software at this time for volume research since I make all my intraday trading decisions based on volume analysis and price action at turning points. There are several issues about deficient real time volume handling in MC. I know I could use older versions, however, I am interested in fixing the beta version of MC so that I can return to using it and Andrew, once convinced, seems very responsive.
Although MC is *THE* top-notch charting program on the market, I am looking forward to the future of MC being the type of software an intraday trader can reliably use to trade with (vis a vis: NT with ATM, or TT) and not just do back-testing research with. It might take a year (or two) more and I would like to support it as much as I can.
Thanks for the suggestions. I use other software at this time for volume research since I make all my intraday trading decisions based on volume analysis and price action at turning points. There are several issues about deficient real time volume handling in MC. I know I could use older versions, however, I am interested in fixing the beta version of MC so that I can return to using it and Andrew, once convinced, seems very responsive.
Although MC is *THE* top-notch charting program on the market, I am looking forward to the future of MC being the type of software an intraday trader can reliably use to trade with (vis a vis: NT with ATM, or TT) and not just do back-testing research with. It might take a year (or two) more and I would like to support it as much as I can.
Exactly my sentiments. It's getting there but MC still has some way to go to become a top notch real-time trading tool. I too would like to give as much support as possible to develop it into the best platform on the market. The charting front-end is supreme but just about everything else about it is behind the competition, and in some cases like the editor and run-time error notification is woefully behind the times. Still it's more than adequate for most of my purposes, and I get around most of the programming deficiencies by writing my own dll's in other IDE's.Although MC is *THE* top-notch charting program on the market, I am looking forward to the future of MC being the type of software an intraday trader can reliably use to trade with (vis a vis: NT with ATM, or TT) and not just do back-testing research with. It might take a year (or two) more and I would like to support it as much as I can.
- RobotMan
- Posts: 375
- Joined: 12 Jul 2006
- Location: Los Altos, California, USA
- Has thanked: 31 times
- Been thanked: 13 times
- Contact:
Hi Super,
Good article, thanks!
I agree with that study. Nothing leads price. What I look for is folks that are "underwater" and the larger traders are squeezing them at possible turns (imaginary places on charts that traders think are important - previous highs and lows, prior day pivots, moving averages, fibs, elliot, gann and other crap). If the markets moves in one direction and orders get filled, the pros will push back and scare them out and collect their stop/exit orders. That's what I am looking for and volume studies will confirm the price action. I am looking for confirmation of retracement or reversal. I am looking for statistically meaningful events (imbalances that shouldn't be there) and always asking "if this is happening, then this *should* happen right here." If it doesn't then I can make a conclusion and not enter. If it does happen then I can make an entry. -always asking "where am I wrong?". -I don't trade price so much as I trade the people that make the price.
Good article, thanks!
I agree with that study. Nothing leads price. What I look for is folks that are "underwater" and the larger traders are squeezing them at possible turns (imaginary places on charts that traders think are important - previous highs and lows, prior day pivots, moving averages, fibs, elliot, gann and other crap). If the markets moves in one direction and orders get filled, the pros will push back and scare them out and collect their stop/exit orders. That's what I am looking for and volume studies will confirm the price action. I am looking for confirmation of retracement or reversal. I am looking for statistically meaningful events (imbalances that shouldn't be there) and always asking "if this is happening, then this *should* happen right here." If it doesn't then I can make a conclusion and not enter. If it does happen then I can make an entry. -always asking "where am I wrong?". -I don't trade price so much as I trade the people that make the price.
- Bruce DeVault
- Posts: 438
- Joined: 19 Jan 2010
- Location: Washington DC
- Been thanked: 2 times
- Contact:
- RichardTodd
- Posts: 12
- Joined: 18 Feb 2010
- Location: Dallas
- Been thanked: 2 times
- Contact:
Using the zen-fire feed, I get huge upticks and downticks that just cannot be correct (more volume in upticks on one bar than ES prints in a day). I'm using MC6beta2.
When I watch the bar in progress, the numbers look reasonable. As soon as a new bar forms, the numbers on the bar that just closed inflate dramatically.
When I watch the bar in progress, the numbers look reasonable. As soon as a new bar forms, the numbers on the bar that just closed inflate dramatically.
- Attachments
-
- upvol_dnvol2.PNG
- screenshot
- (166.99 KiB) Downloaded 1816 times
- RobotMan
- Posts: 375
- Joined: 12 Jul 2006
- Location: Los Altos, California, USA
- Has thanked: 31 times
- Been thanked: 13 times
- Contact:
Hi Richard,Using the zen-fire feed, I get huge upticks and downticks that just cannot be correct
Yes, several of us have noticed this. I know that you and Bill use range bars quite a bit at EOTpro and it is unfortunate that you are unable to use MultiCharts because of this and because of the intrabar persist problem. I am in the same boat and am forced to rely on NT for research at this time as TS is also insufficient for my needs.
I don't know what is happening with MC development. Andrew (or anyone at TSS) has not responded to any of the questions asked in the past several weeks. It is quite disconcerting. Many other very important questions about MC and PLE have been asked that only TSS can answer, but so far none have been forthcoming. I hope that all is well on Andrew's end and that someone is able to attend to the needs of this neglected forum soon.
- Bruce DeVault
- Posts: 438
- Joined: 19 Jan 2010
- Location: Washington DC
- Been thanked: 2 times
- Contact:
- RobotMan
- Posts: 375
- Joined: 12 Jul 2006
- Location: Los Altos, California, USA
- Has thanked: 31 times
- Been thanked: 13 times
- Contact:
You know this for a fact? I mean, for several years they have been aggressively working but always found time to answer and clarify issues raised by this forum's users. Now, all of a sudden, not a peep. Have you been in contact with Andrew? Did he say, "Gosh, I just don't have an hour of my time to answer questions in the Multicharts forum for the loyal users of my product because I am coding so much more aggressively than in the past. Hell, I don't even have time to just drop in for a second to let them know I am so busy that I can't respond right now."They are aggressively working to finish up 6.0. It seems likely they will be more active once that is completed and the next steps have been taken.
No, my friend, the reason for this neglect is something other than aggressive coding.
- Bruce DeVault
- Posts: 438
- Joined: 19 Jan 2010
- Location: Washington DC
- Been thanked: 2 times
- Contact:
- RichardTodd
- Posts: 12
- Joined: 18 Feb 2010
- Location: Dallas
- Been thanked: 2 times
- Contact:
Just as a point of clarification, I'm not affiliated with EOTPro anymore. But, I do still use range bars a lot, true enough!Yes, several of us have noticed this. I know that you and Bill use range bars quite a bit at EOTpro and it is unfortunate that you are unable to use MultiCharts because of this and because of the intrabar persist problem.
It's lucky that the numbers are ridiculous... if they were just a little off I'd probably have trusted them! I'm sure the issue will be addressed by the 6.0 release or shortly afterwards, as basic volume reporting is pretty fundamental to a professional platform.
- Dave Masalov
- Posts: 1712
- Joined: 16 Apr 2010
- Has thanked: 51 times
- Been thanked: 489 times