Posted: Tue Jun 06, 2006 6:17 am
AFJ
I was a bit lazy and usually posted images straight out of screen capture without re-sizing. Thanks for reminding, done that.
My main point regarding my post above is if using fixed dates, we have to actually investigate carefully because the back-test result may not be as it seems. If for example, during rollover with a large spread and low volume, the back-tested price may be far from the actual fill-price. Hence the commission/slippage we set in TBB may be un-realistic. Also, depending on how conservative/aggressive we set the "minimum futures volume" in TBB, this change in volume could actually mean taking/not taking the signal in back-testing - as least as I understand it.
I was a bit lazy and usually posted images straight out of screen capture without re-sizing. Thanks for reminding, done that.
My main point regarding my post above is if using fixed dates, we have to actually investigate carefully because the back-test result may not be as it seems. If for example, during rollover with a large spread and low volume, the back-tested price may be far from the actual fill-price. Hence the commission/slippage we set in TBB may be un-realistic. Also, depending on how conservative/aggressive we set the "minimum futures volume" in TBB, this change in volume could actually mean taking/not taking the signal in back-testing - as least as I understand it.