Bending the rules

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
Jwebster
Roundtable Fellow
Roundtable Fellow
Posts: 52
Joined: Fri Jul 04, 2003 10:23 am
Location: London, UK

Bending the rules

Post by Jwebster »

Here's an interesting one (I hope). My Hang Seng index trading system had me short most of March but gave me an exit signal yesterday given the rising market. As my system is end of day I would have exited at today's open. Ok, but here's where it gets tricky. Looking at the US market last night I saw it was down significantly on interest rate fears and I knew that the Hang Seng would follow. Given that I was still short I could leave my position on and ride today's downfall (-361 points, nice!) or exit as my system told me to at the open (still 100-odd points down from yesterday's close). So what would you do? Override the system because you 'know' the market's going to get hammered, or be a good boy and just take the signal and accept a smaller profit. Am i just being greedy or what? Any thoughts appreciated.
RangeRover
Contributing Member
Contributing Member
Posts: 8
Joined: Fri Jun 06, 2003 8:57 pm

Post by RangeRover »

Unfortunately, these may be closer to feature than process, but here goes anyway:

I would like to be able to compare the profitability of an entry method to the profitability of random entry. This might require consideration of all possible choices for entry, otherwise every random entry run would be different. This would help me put into proper perspective the worth(lessness) of any entry method.

It would be nice to easily be able to port common types of analysis software code to the testing platform, i.e. can I pipe Matlab m-files, various calculus functions, a finite difference subroutine from FORTRAN, etc. to the software. While one can almost always convert one format to another, it would be nice to not be compelled to do so. This way one could use a third party software for it's area of strength, and use the testing platform to test.

Ideally, once a system is designed, the platform should be able to implement every aspect of trading, without human intervention, though human intervention should still be allowed. I know, you shouldn't override your system, but there are still manual overrides for aviation autopilots. I'm certain I cannot forsee all possible combinations of events; I might have to pull the plug!

I hope these aren't too far from what you were looking for.

RR
Post Reply