[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Systems Criteria



PureBytes Links

Trading Reference Links


In a message dated 9/16/98 8:03:42 AM, magnus@xxxxxxxxxxx writes:

<<Can some of the more experience systems developer answer a few questions?



Of all the criteria on the systems report in TS, what would you consider as

an important  criteria in developing a trading system? And what would you

consider as acceptable numbers for the following:



1) Percentage of winning trades?

2) Ratio of Avg. Win/Loss?

3) Profit Factor?

4) Ratio Largest Win/Largest Loss?

>>

As far as important criteria I would rate a tie between profit factor and
percentage of winning trades.  Regarding acceptable numbers, it can vary
depending on the type of system but here are some numbers just to contribute
to the discussion.

1) 45% or more winning trades.  A high percentage of winners is important to
those who operate systems in real time.  If odds are in your favor you are
less likely to second guess the system and skip trades.  If you lose most of
the time it makes you want to skip trades so you skip a loser or two and then
you miss the big winner.  Usually that is when good systems are abandoned and
the developer is blemed for marketing trash.
2) Ratio of average win to avg loss can vary depending on winning percentage
but should be at least 1.0 for most systems
3)  Profit factor seems to get smaller with activity.  On longer term systems
it should be about 2.5 or higher.  On shorter term systems with appropriate
commissions and slippage I would settle for anything in the neighborhood of
1.5 to 2.0
4) Ratio of largest win to largest loss - I don't like this ratio and only
look at it in the reverse sense.  Seeking these rare mega trades can lead to
over optimization and other problems.  If the largest winner is a large
percentage of the profits the system may not be robust as largest winner may
not occur again.  I usually throw out the largest winner or two and take a
second look at the results and ratios assuming that we may not catch those
trades in the future.  How many times is sugar going to go from 15 cents to 60
cents in our lifetime?

Good question. Hope to see other responses as I just scratched the surface.

Chuck
traderclub.com