A factor that people keep expressing for what constitutes a good
strategy is that the strategy should show consistent profits over the
backtest period (i.e. $100/day over 10 days instead of $1000 on one
day and flat the rest).

In other words, the profit curve should be "smooth" and, ideally,
steep.  Unfortunately, there's no easy way to optimize for this in the
current implementation.  I propose that we could measure these factors
by performing a linear regression across the set of trades for a given
strategy and measuring the standard deviation of the trades from the
resultant line-of-best-fit.  A strategy that has a steep, positive
line-of-best-fit, as well as a low standard deviation would be an
ideal candidate for trading because it would have been highly
profitable over the entire backtest period.

The downside of incorporating this sort of calculation into the
optimizer is that it would add a lot of overhead into an already CPU-
intensive process.  However, it could also provide a lot of insight
into what constitutes a good strategy.

Thoughts?
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"JBookTrader" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/jbooktrader?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to