Your Cumulative delta chart type is giving incorrect results. Are you aware of this?
I've compared it against 2 other programs that both give same results with same data.
Also, I can't use Cumulative delta chart type anymore because it slows the response of Multicharts so much that I can't use it anymore. The lag is terrible and so I can't tolerate so much lag.
Regards
Cum Delta
-
Henry MultiСharts
- Posts: 9165
- Joined: 25 Aug 2011
- Has thanked: 1264 times
- Been thanked: 2957 times
Re: Cum Delta
Hello dtl-saw,
Please describe the exact issue you have and attach screenshots demonstrating it.
Please describe the exact issue you have and attach screenshots demonstrating it.
Re: Cum Delta
I'm using MultiCharts .NET64 Version 9.1 Release (Build 11817)
The specific problem is that the cumulative delta "ask traded vs bid traded" does not give proper shape when compared with 2 other golden standards: 1) InvesterRT 2) SC
I've attached 3 different screenshots of price and cumulative delta from 3 different charting programs on 10/21 day session of ES (S&P500-emini December 2015 contract).
The 3 programs are:
1) Multicharts from my computer using CQG data.
2) SC from my computer using CQG data.
3) InvesterRT from FT71 youtube video using IQfeed data.
In all screenshots the instrument price is on top panel and the cumulative delta value is on lower panel.
The InvesterRT screenshot shows a completely independent version of the data as this came from a youtube video on FT71 trader bite (here's the link https://www.youtube.com/watch?v=T9ZiELaV0dg). He has publicly stated that he only uses IQFeed for data and this screenshot is from 10/22 trader bite video during a timeframe when he is reviewing the previous days price action on 10/21.
Now referring to the 4 specific data points annotated on Multicharts and SC:
1.) from price #1 to price #2
a.) Multicharts cumulative delta drops by over 40,000 or from 915 to -37,560
b.) SC and InvesterRT cumulative delta drops by less than 20,000 or from -2136 to -20,031
The absolute differences in cumulative delta wouldn't be difficult to deal with as long as Multicharts was consistent throughout the day, but it's not!
2.) from price #2 to price #3
a.) Multicharts cumulative delta drops by about 10,000 or from -37,560 to -48,000
b.) SC and InvesterRT cumulative delta drops by almost 30,000 or from -20,031 to -51,911
3.) at the end of day (EOD) just comparing against a very recent swing low not more than 20 minutes before at Price #3
a.) Multicharts cumulative delta is higher than that at Price #3
b.) SC and InvesterRT cumulative delta is lower than at Price #3
When Multicharts first introduced cumulative delta I performed this same comparison of your program and concluded that your cumulative delta routine was good. This is the first time since then that I've done the comparison and I think this function is now broken.
There are many days were Multicharts cumulative delta looks very close to the other golden standards. But there are days like 10/21/2015 that are just completely broken and for that reason Multicharts cannot be trusted for this type of data.
I just realized that this post should have gone in the "Project Management" section so if you can move this post there please do.
The specific problem is that the cumulative delta "ask traded vs bid traded" does not give proper shape when compared with 2 other golden standards: 1) InvesterRT 2) SC
I've attached 3 different screenshots of price and cumulative delta from 3 different charting programs on 10/21 day session of ES (S&P500-emini December 2015 contract).
The 3 programs are:
1) Multicharts from my computer using CQG data.
2) SC from my computer using CQG data.
3) InvesterRT from FT71 youtube video using IQfeed data.
In all screenshots the instrument price is on top panel and the cumulative delta value is on lower panel.
The InvesterRT screenshot shows a completely independent version of the data as this came from a youtube video on FT71 trader bite (here's the link https://www.youtube.com/watch?v=T9ZiELaV0dg). He has publicly stated that he only uses IQFeed for data and this screenshot is from 10/22 trader bite video during a timeframe when he is reviewing the previous days price action on 10/21.
Now referring to the 4 specific data points annotated on Multicharts and SC:
1.) from price #1 to price #2
a.) Multicharts cumulative delta drops by over 40,000 or from 915 to -37,560
b.) SC and InvesterRT cumulative delta drops by less than 20,000 or from -2136 to -20,031
The absolute differences in cumulative delta wouldn't be difficult to deal with as long as Multicharts was consistent throughout the day, but it's not!
2.) from price #2 to price #3
a.) Multicharts cumulative delta drops by about 10,000 or from -37,560 to -48,000
b.) SC and InvesterRT cumulative delta drops by almost 30,000 or from -20,031 to -51,911
3.) at the end of day (EOD) just comparing against a very recent swing low not more than 20 minutes before at Price #3
a.) Multicharts cumulative delta is higher than that at Price #3
b.) SC and InvesterRT cumulative delta is lower than at Price #3
When Multicharts first introduced cumulative delta I performed this same comparison of your program and concluded that your cumulative delta routine was good. This is the first time since then that I've done the comparison and I think this function is now broken.
There are many days were Multicharts cumulative delta looks very close to the other golden standards. But there are days like 10/21/2015 that are just completely broken and for that reason Multicharts cannot be trusted for this type of data.
I just realized that this post should have gone in the "Project Management" section so if you can move this post there please do.
- Attachments
-
- RT2.png
- InvesterRT data from 10/21/2015
- (858.12 KiB) Downloaded 1421 times
-
- SC.png
- SC data from 10/21/2015
- (80.52 KiB) Downloaded 1393 times
-
- Multicharts.png
- Multicharts.Net data from 10/21/2015
- (69.27 KiB) Downloaded 1418 times
Re: Cum Delta
Yes I have noticed issues with this tool as well but I'm unable to rule out datasource issues, in my case. Something to consider is that real-time vs historical data will be different and as a result will of course change the shape. You need to compare real-time with real-time and vice versa.
Esignal Data feed comment:
As an aside, esignal have acknowledged a data source issue on HKFE futures. I have been chasing them about this since June 2015 and escalating the issue all the way through to their 'VP of Global Support' Scott Johnson. I sadly must report that their support is the worst I have ever experienced. They have acknowledged data discrepencies, but failed to communicate back to me, or notify any other customers. The worst of it is that they have just done nothing about it. Happy to keep taking my money for a flawed service. I cannot recommend esignal to anyone unless they really need them. The only reason I am there is they are the only ones offering global futures data at a price that doesn't cost and arm and leg.
With kind regards,
MK
Esignal Data feed comment:
As an aside, esignal have acknowledged a data source issue on HKFE futures. I have been chasing them about this since June 2015 and escalating the issue all the way through to their 'VP of Global Support' Scott Johnson. I sadly must report that their support is the worst I have ever experienced. They have acknowledged data discrepencies, but failed to communicate back to me, or notify any other customers. The worst of it is that they have just done nothing about it. Happy to keep taking my money for a flawed service. I cannot recommend esignal to anyone unless they really need them. The only reason I am there is they are the only ones offering global futures data at a price that doesn't cost and arm and leg.
With kind regards,
MK
-
Henry MultiСharts
- Posts: 9165
- Joined: 25 Aug 2011
- Has thanked: 1264 times
- Been thanked: 2957 times
Re: Cum Delta
Hello dtl-saw,
Thank you for the detailed description. Our engineers have studied your case.
We have compared the CumulativeDelta chart using IQfeed data in MultiCharts with the screenshot you have provided (RT2.png) and have not found any difference.
As for the SC.png - SC and MultiCharts are getting CQG historical data from different sources, therefore that is not relevant to compare the charts built using different source data. In MultiCharts .NET 9.1 Release 1 there is also a known issue with the CumulativeDelta chart - it is being plotted incorrectly after changing the Data Range value of the chart. This issue will be fixed in ver. 9.1 Release 2.
Thank you for the detailed description. Our engineers have studied your case.
We have compared the CumulativeDelta chart using IQfeed data in MultiCharts with the screenshot you have provided (RT2.png) and have not found any difference.
As for the SC.png - SC and MultiCharts are getting CQG historical data from different sources, therefore that is not relevant to compare the charts built using different source data. In MultiCharts .NET 9.1 Release 1 there is also a known issue with the CumulativeDelta chart - it is being plotted incorrectly after changing the Data Range value of the chart. This issue will be fixed in ver. 9.1 Release 2.