Portfolio Filter inflection point and trades

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
Post Reply
MarkS
Roundtable Fellow
Roundtable Fellow
Posts: 70
Joined: Wed Dec 15, 2010 12:10 pm
Location: Chicago, IL

Portfolio Filter inflection point and trades

Post by MarkS » Fri Jun 17, 2011 3:23 pm

I noticed the last few days that my MACD filter on the Donchian system is at an inflection point on one commodity I trade. So it will show long trade entries for a few days, then the next day the long disappears (it is rejected) and the order list shows a short trade entry point as the MACD flips about.

Yesterday, using data from 6/15, I had a buy stop order to go long this market on 6/16. This did not get triggered. However, once I downloaded the data after the close on 6/16, it no longer showed any active long order (it shows up on the rejected list spreadsheet instead) but it does show that a short order was supposed to be filled.

Going back to my generated orders on 6/15, there was no indication to have a short stop in the market at all, since it was being rejected by the portfolio filter and I was supposed to have a long order active. Anyone else had issues similar to this? Is there a reason why TB is assuming it went short? Any workarounds?

LeviF
Roundtable Knight
Roundtable Knight
Posts: 1432
Joined: Mon Dec 22, 2003 12:24 pm
Location: Des Moines, IA
Contact:

Post by LeviF » Fri Jun 17, 2011 3:49 pm

Maybe another case of change in the data? What happens now if you set test end date to 6/15 and generate orders?

Tim Arnold
Site Admin
Site Admin
Posts: 8767
Joined: Tue Apr 06, 2004 1:41 pm
Location: Boston, MA
Contact:

Post by Tim Arnold » Fri Jun 17, 2011 4:26 pm

Set the test end date, generate orders, then set the test end date ahead and generate orders. If everything looks ok then it was a data change issue. One of the things I do is use Generate Forward in CSI UA to help stabilize the data. But if CSI made data corrections or otherwise, then things can change as the historical data changes.

MarkS
Roundtable Fellow
Roundtable Fellow
Posts: 70
Joined: Wed Dec 15, 2010 12:10 pm
Location: Chicago, IL

Post by MarkS » Fri Jun 17, 2011 8:08 pm

Thank you gentlemen. Tim -- That is exactly what I had done to test the order, and everything else across all markets was fine, thus it must be a data issue. The contract in particular is the Rough Rice, July. Disturbing to see how data can "change" after the fact. Now the decision becomes to chase the order 1.6%, let it slide, or put in a limit and see if I get filled (but then it is technically moving in the wrong direction!).

marriot
Roundtable Knight
Roundtable Knight
Posts: 365
Joined: Thu Nov 20, 2008 3:02 am

Post by marriot » Sat Jun 18, 2011 1:27 am

Hi Tim,
what could be the problems if you make TB to save data allowing Csi just
to add Last day?

thank you.

AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2075
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner » Sat Jun 18, 2011 4:36 am

MarkS wrote:thus it must be a data issue
Each day make a copy of all your futures and forex ASCII files and place them in a separate dated folder for each day. When you come across such issues you can then easily track down data differences. Once a month or so you can then delete month old ASCII files so as not to accumulate too much clutter. LINK

michaelt
Senior Member
Senior Member
Posts: 37
Joined: Wed Feb 03, 2010 10:53 am
Location: princeton,nj

Post by michaelt » Mon Jun 20, 2011 10:57 pm

Hi AFJ, I'm interested in this discussion and your and Tim's suggestions - I see the confusing impact of CSI historical adjustments. You posted a link but for me it's empty.

Moderator's note: that link points into the Trading Blox Customers Only area of the website. It is not accessible unless you're a customer ---> (tutorial)

Post Reply