Announcement

Collapse
No announcement yet.

Back Test "Trades" Report for identical strategy reporting different results?

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Back Test "Trades" Report for identical strategy reporting different results?

    I have the same EFS strategy running in a chart that is utilizing external time intervals. When the strategy is loaded into two different charts (same symbol) with the chart interval of one set to the smallest interval accessed in the EFS strategy I get one set of trading results. When I change the chart interval to a 1 minute chart the back test reports reports are dramatically different.

    See the trade reported in the first image below. The long trade is reported at 9:57 with an entry price of 7542.75 and closed at 10:07 at an exit price of 7543.25 for a net loss of $4.00. This trade report is derived from the chart with the chart interval of 1 minute.

    Click image for larger version  Name:	backtesterror1.png Views:	2 Size:	10.2 KB ID:	268310

    The second image shows the trade report for the identical strategy EFS loaded into the same Symbol NQ M19, the only difference is the chart interval for the second is changed to the same time interval as the smallest time interval accessed within the chart. This long trade also reported at 9:57 has an entry price of 7551.00 and closed also at 10:07 at an exit price of 7540.25 for a net loss of $229.00. That is a inaccuracy/difference of over 10 full points in the NQ futures contract.
    Click image for larger version  Name:	backtesterror2.png Views:	2 Size:	10.5 KB ID:	268311

    Given that different prices are reported at the exactly the same minute in the report proves that one and perhaps both of the above reports are inaccurate?

    Frequently when reporting software issues defining a code "defect" can be an ambiguous problem. In this case there is no ambiguity, as both reports reporting the trades opening and closing price at exactly the same time intervals cannot be correct, one is apparently inaccurate.

    This strategy when run real time executes trades at entirely different times and prices than those reported in either of the above reports. This live trading results are drastically worse then either of the above back testing reports.

    My questions are does eSignal have any plans of 1) acknowledging the above discrepancies and more importantly 2) Plan to correct the back testing inaccuracies which can lead to large losses should one depend on the incorrect reports for actual trading.

    I hope this helps others avoid the emotional and financial damage incurred by basing trading decisions on eSignals back testing reports which IMHO is the best product overall in the business, but not without problems as described above.

    What's very disappointing is in the past (I have been a user for 15+ years) Avery, Alexis, JasonK, etc. would have been very much engaged in generously applying their expertise in addressing this type of issue. Apparently after the more recent corporate changes occurred the resources of the past are no longer made available and that cannot be in the long term interest of either eSignal or it's dedicated user base.

    In the past this forum community was very active and the interaction between the community and eSignal staff was invaluable to me in becoming more successful as a trader. I encourage anyone with similar trading goals to comment/participate as typically there were dozens of forums entries daily and now weeks go by without any activity at all. This cannot be in the best interest of the community or perhaps more importantly the future viability of eSignal as a product.

    Hope this helps.

    glen

    glendemarco@gmail.com
    Attached Files
    Last edited by demarcog; 05-19-2019, 08:27 AM.
    Glen Demarco
    glendemarco@gmail.com

  • #2
    I reported this issue with the back tester two months ago and just checking in to see if any users or eSignal folks have had a chance to review this issue?

    I am assuming it is an error or "bug" and thought eSignal folks might have weighed in. I make the assumption based on the fact that the same strategy EFS loaded on the same symbol produces two drastically different trade results on historical data. One shows a very successful strategy and the other does not. In real time the results are much less then.

    The only difference being the chart interval of the two charts and since both results are vastly different I am assuming one is a bug, or incorrect.

    I also opened an incident with support on this issue and would very much appreciate an update on the issue.

    I can't over emphasize how much users such as myself rely on the relatively accuracy of the back tester.

    Any assistance on how to properly utilize the back testing tool when either the chart interval is different from the EFS (ie, ema(20,inv(15)) on a 5 minute chart or when multiple time intervals are utilized in the EFS inv(15) & inv(60) etc would be extremely helpful to some of us.

    So if anyone can suggest a solution please do.



    thank you very much.

    glendemarco@gmail.com
    Last edited by demarcog; 07-12-2019, 07:30 AM.
    Glen Demarco
    glendemarco@gmail.com

    Comment

    Working...
    X