Support Board
Date/Time: Sun, 12 Jan 2025 09:15:19 +0000
Backtesting in 3-line break or BetterRenko charts
View Count: 3821
| [2017-02-08 12:10:20] |
| Orion - Posts: 12 |
|
Hi SC Support or users, The backtesting results of some "non-natively-supported" chart types like 3-line break or BetterRenko seem to be inaccurate. Is that an existing limitation or something that I have not done properly? Please kindly advise and sorry if I have overlooked anything. Thank you. Regards, Orion Date Time Of Last Edit: 2017-02-08 18:06:37
|
| [2017-02-08 18:25:43] |
| |
|
We cannot answer this question and it is not within the scope of our support. Refer to: Automated Trading Management: Troubleshooting Automated Trading System Behavior Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing Date Time Of Last Edit: 2017-02-08 18:26:18
|
| [2017-02-08 19:55:16] |
| Orion - Posts: 12 |
|
Thank you for your prompt response. Perhaps I haven't stated my question clear enough. Attached please find two pictures of a 5-min crude oil (CL) chart: 1. The first one has only the BetterRenko study applied, with all input parameters being the default except that "Show real open" was set to "Yes". There is no problem in this picture. 2. The second one has the BetterRenko and the "Trading Example" (TargetValue=0.3; Stop Value = 0.1) studies applied. The image was captured right after a Bar-Based BackTest. As shown in the second picture, the chart was somehow "corrupted" and the entries and exits of trades are clearly not correct. Since both studies are bundled with SC, I guess you might be able to reproduce the result. Apart from BetterRenko, the 3-line break study exhibits similar behaviors as well. Hope this would better describe my question. Any hints would be highly appreciated. Thank you. Date Time Of Last Edit: 2017-02-08 20:12:06
|
| [2017-02-08 20:33:37] |
| |
|
Better Renko is a user contributed study and we provide absolutely no help with it. We are removing it from the list of User Contributed studies. Do not use it. Follow the instructions here to use the Renko Bar Period Type: Renko Bar Charts If you are using this study: https://www.sierrachart.com/index.php?page=doc/TechnicalStudiesReference.html#s131 It will be most accurate to perform a replay back test and not a bar based back test. Although even with the bar based back test, we only noticed minor inconsistencies compared to when the chart bars are built during a full recalculation. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing Date Time Of Last Edit: 2017-02-08 20:34:48
|
| [2017-02-08 21:11:11] |
| Orion - Posts: 12 |
|
Thank you. It is clear now. Yes, I am using that "standard" 3-line break study, and as you have mentioned, it does function correctly in a replay backtest. However, if a bar-based backtest is performed, the same occured - the chart was "corrupted" and the entries and exits of trades are not correct. Please see the attached pictures. Could that be fixed (or have I overlooked anything)? The bar-based backtest is far more convenient than replay for system prototyping. On a side note, BetterRenko IMHO is a useful study. I suggest keeping it (no matter it can be back-tested correctly or not) unless there is a similar alternative. Thanks for your consideration! Date Time Of Last Edit: 2017-02-08 21:14:50
|
 Before_Backtest.jpg / V - Attached On 2017-02-08 21:09:16 UTC - Size: 256.85 KB - 438 views After_Backtest.jpg / V - Attached On 2017-02-08 21:09:24 UTC - Size: 294.67 KB - 414 views |
| [2017-02-08 21:18:20] |
| |
|
Could that be fixed
No.We also do not observe the result that you do in those chart images. Although we tested with 1 minute underlying bars. When posting an image of a chart use only this method: https://www.sierrachart.com/index.php?page=PostingInformation.php#Image Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing |
| [2017-02-08 21:43:39] |
| Orion - Posts: 12 |
|
I have repeated the test with 1-minute underlying bars and the result is the same: http://www.sierrachart.com/image.php?Image=1486589361707.png Any ideas? Please advise if this is an issue which you have no plan to fix, or it happens only at my side. Thanks. Date Time Of Last Edit: 2017-02-08 21:45:00
|
| [2017-02-08 21:59:39] |
| |
|
This is not what we see at all. And we also see the word "Replay" at the top left of the chart indicating you are actually running a chart replay. But that should not matter anyway. Explain to us exactly how you ran the back test. We are not aware that there is any issue to fix. The slight discrepancy we saw between a bar based back test and a full recalculation of these Three Line Break chart bars has a certain reason why it occurs and it is only very small. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing Date Time Of Last Edit: 2017-02-08 22:07:04
|
| [2017-02-08 22:26:39] |
| Orion - Posts: 12 |
|
Thanks for your help. Here are what I did - I start from scratch here and am repeating the steps while typing this message: 1. Open a new intraday chart ("CL-201703-NYMEX"). In the Chart Settings, set "Days to load" to 10, and "Graph Draw Type" to Candlestick Bars. 2. Enable Trade->Show Order Fills. 3. Add the "Three Line Break Chart" study, and set "Show Real Open" to "Yes". This would lead to the following clean chart: http://www.sierrachart.com/image.php?Image=1486591617543.png 4. Add the "Trading Example" study, set "Enabled" to yes, "Target Value" to 1 and "Stop Value" to 0.5. 5. Perform Trade->"AutoTrade System Bar Based BackTest". The following chart is resulted: http://www.sierrachart.com/image.php?Image=1486592190696.png 6. Now, I perform Trade->"AutoTrade System Replay BackTest" (which took much longer), everything becomes normal again: http://www.sierrachart.com/image.php?Image=1486592600502.png Not sure why this happened. Thanks again. |
| [2017-02-09 17:46:09] |
| |
|
We do see the problem now following these instructions. We will be looking into this, but it is quite involved and we do not expect to get to it for many months. You need to use a replay back test and not a bar based back test. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing |
| [2017-02-09 18:04:58] |
| Orion - Posts: 12 |
|
Hi SC Support, I understand, thank you for looking into it. Meanwhile, since your 1-min chart was normal for a bar-based test, may I know your previous chart settings? I just wonder whether the problem can somehow be bypassed (or, which of my steps have "triggered" the problem?). Any hints would be highly appreciated. Thanks! |
| [2017-02-11 05:49:14] |
| |
|
The only difference was this: "Show Real Open" to "Yes". We have not spent further time to do further testing and analyze what the reason for the difference is. We will be looking into this later on when we can. Sierra Chart Support - Engineering Level Your definitive source for support. Other responses are from users. Try to keep your questions brief and to the point. Be aware of support policy: https://www.sierrachart.com/index.php?l=PostingInformation.php#GeneralInformation For the most reliable, advanced, and zero cost futures order routing, *change* to the Teton service: Sierra Chart Teton Futures Order Routing |
To post a message in this thread, you need to log in with your Sierra Chart account: