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

RE: is profit factor the best measure



PureBytes Links

Trading Reference Links

Richard, thanks for the reply.

I think the equity curve might be a level deeper than I am looking at the 
moment - it requires trade by trade details.

Lets say I have a new trading idea I want to test, and it has 4 parameters.
I do an initial optimisation run of 1600 tests using a wide sweep of the 
parameter values to get a feel for where it might operate best.   When I 
look at the results of the optimisation, I am trying to narrow the 
parameters down before looking in more details at trade-by-trade details.

My question arose because I did the above and got a very consistent set of 
profit factors - all were between 1.4 and 2.1 - not spectacular, but 
consistent for a first cut.   The net profits were widely separated, but 
generally in line with the number of trades. (between 200 and 300 trades) 
The max drawdowns were fairly consistent.

So, which ones to choose to work on further ??

Rgds
Peter

-----Original Message-----
From:	Richard [SMTP:rjb@xxxxxxxxxxxxxxxxxxx]
Sent:	Saturday, August 22, 1998 1:09 AM
To:	RealTraders Discussion Group
Subject:	Re: is profit factor the best measure

Nothing wrong with it, but I think the best thing to look at is the equity
curve. Was the ride up a smooth one or a rough one? And I also want to be
sure that superior performance for, say a ten year test, wasn't based on
incredible performance in year one and two and mediocre or poor performance
in the rest of the period. After all, would you want to trade a system that
did well overall but barely broke even for many years at a stretch?

-RB

>Is profit factor (gross profit / gross loss) the best measure to compare
>systems ?
>
>I have tested several systems and often the number of trades differs
>dramatically with different parameter sets.
>This can have a huge affect of the net profit of different combinations,
>and seems to make straight profit meaningless as a comparison measure.
>
>I notice that some journals do not even list profit factor as part of 
their
>results.
>
>Is this something wrong with profit factor ?
>
>All comments appreciated.
>
>(I use tradestation)
>
>Thanks
>Peter