|
PureBytes Links
Trading Reference Links
|
The problem is "bouncing ticks" - see also online help/manual
In realtime TS "knows" the sequence of ticks coming in to form a bar, i.e. was
the high reached first or the low - in historical testing it ASSUMES it, therefore
guessing it wrong one time or the other.
So it might hit a stop or profit target first - besides in realtime it was different
Best way around is to use as short as possible time bars, i.e. try to CHANGE
the code of your system to run on 1min bars instead on 5min or 60min - often
thats possible, but not allways.
.....and its not a bug - its a "feature" <g>
rgds hans
> I'm fairly sure I remember this being discussed before, but
> I can't find previous references and I don't remember what the
> solution (if any) was. I believe the example that was described
> before was an over-a-weekend case, which isn't what I'm seeing.
>
> I'm following a system in realtime. Yesterday it ended a long
> position with a reversal, selling into a short position.
> Last night I rebooted my system. When I restarted Tradestation,
> surprise! The reversal sell turned into a simple exit with
> no current short position.
>
> In other words, the results that the system shows on historical
> data (yesterday) are different than what it showed me in realtime.
>
> Obviously this is a serious concern if you trust backtested systems.
> If the realtime behavior is different than the backtested behavior....
>
> Is this an insidious TS bug, or what? Is there some way I can avoid it??
>
> Thanks,
> Gary
>
|