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

[EquisMetaStock Group] Re: Opinion about Backtesting and results



PureBytes Links

Trading Reference Links

------------------------------------------------------------------------
You cannot reply to this message via email because you have chosen not
to disclose your email address to the group.

To reply:  http://groups.yahoo.com/group/equismetastock/post?act=reply&messageNum=6129
------------------------------------------------------------------------

Thanks a lot for your answer Dan.

Other question for you: Do you use, in addition to your system, 
tradition tecnical analisys based on grafic formation ?

Only a personal comment: you said that you compare system over the 
same data, I must think that is the same of running the systems 
against a single stock. Maybe is my system, but I've got just 
disapointed when my beautiful system gave me only good results for 
few stocks and large bad results for the most stocks ... this is 
something to think about, at least I will ...

Best Regards,
Marcio.

--- In equismetastock@xxxxxxxxxxxxxxx, "Dan Harels" <harelsdb@xxxx> 
wrote:
> > > Some feedback about results would be very nice ... which generic
> > > numbers would be considered "excelent", "good" or "acceptable" 
(or
> > > any aditional criteria you want) in:
> > > Average Profit per month: ???
> > > Winnin/Losing trades: ??/??
> > > AvgWin/AvgLoss
> > >
> > > Thanks a lot.
> >
> I don't know the short answer to your question.  Two possibilities 
are; it 
> depends or it is relative.
> 
> A system should be judged based on expectancy.  Expectancy is 
simply the 
> amount of profit that a system can extract from a given security 
over a 
> selected period of time.  Expectancy is a function of the number of 
trades 
> signalled, the percentage of those trades that are winners and the 
size of 
> the winners relative to loosers.  All three factors are important 
and a 
> system cannot be judged without considering all of them.  For 
example, a 
> system might have a high percentage of winners and winners that are 
much 
> larger than loosers but have a low expectancy because it triggers a 
small 
> number of trades.  On the other hand, another system might have 
only small 
> advantages in terms of percentage of winners and relative size of 
winners to 
> loosers but have a high expectancy because the system triggers a 
very large 
> number of trades.  Las Vegas was built on the second type of system.
> 
> Personally, I calibrate my systems on one data set and test them 
against 
> another.  My calibration set is very large and I am fairly 
confident that I 
> am not over-optimizing.  I test all of my systems against the same 
data so I 
> can compare my bottom lines on each system and judge their relative 
> performance.
> 
> Just me,
> 
> Dan
> 
> _________________________________________________________________
> Protect your PC - get McAfee.com VirusScan Online  
> http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963


To unsubscribe from this group, send an email to:
equismetastock-unsubscribe@xxxxxxxxxxxxxxx

 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/