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

[amibroker] Re: Real World Systems



PureBytes Links

Trading Reference Links

Are you sure about that ?  How are the trend lines drawn ?

--- In amibroker@xxxxxxxxxxxxxxx, "epintoem" <epintoem@xxxx> wrote:
> A real world system that is profitable is one that is based on the
> dynamics of market movement. Indicators are a wonderful tool but 
they
> fail at times because they do not take into account "reading the
> tape". I use short term trend lines to trade futures and for the 
most
> part if the system is followed can be profitable depending on the 
time
> frame one wants to trade. Unfortunately, because of the way the 
trend
> lines are drawn there is no mechanical way to test it.
> 
> 
> --- In amibroker@xxxxxxxxxxxxxxx, Fred Tonetti <ftonetti@xxxx> 
wrote:
> > Ken,
> > 
> > Although I don't disagree with your list of "bad ones" per se I
> > characterize and look at these somewhat differently ...
> > 
> > 1.	Too narrow a market timeframe test so as not to cover both 
bull
> > and bear periods.  IMHO those that adopt the practice of writing 
systems
> > to test only bullish periods or only bearish periods are asking 
for
> > trouble.
> > 
> > 2.	Too few signals for the results to be statistically 
significant.
> > Personally I like to have systems that show lots (read hundreds) 
of
> > signals that occur end to end NOT side by side.  What I mean by 
that is
> > if I have a system that potentially trades 100 different things ( 
Like
> > the individual stocks of NDX ) and over the course of time they 
generate
> > 1000 trades but lots of those trades overlap in time then to my 
way of
> > thinking these overlapping trades do not constitute end to end 
trades
> > and therefore do not count.
> > 
> > 3.	As far as the remainder of your list goes, to me this all 
comes
> > down to the robustness of the system which can mean several 
things but
> > the simplified explanation to me means it either does well on out 
of
> > sample or it doesn't.  I don't particularly care if a system has 
too
> > many variables ( what is too many ? 3 ? 30 ? 300 ? ) or the 
increment is
> > too small ( what is too small ? 5 ? 0.5 ? 0.05 ? ) or the system 
can be
> > used only on one segment of the market(s) i.e. I'd be quite happy 
having
> > a system that I could trade S&P e-mini's with if it worked well 
in & out
> > of sample even if it couldn't be used for any other thing.
> > 
> > Fred
> > 
> > > A "bad one has....
> > > ....too many variables
> > > ....too short a time period
> > > ....to small an increment in the optimizing variable
> > > ....too few issues/too many issues
> > > ....each stock/or future has its own optimized parameters


------------------------ Yahoo! Groups Sponsor ---------------------~-->
Get a FREE REFINANCE QUOTE - click here!
http://us.click.yahoo.com/2CXtTB/ca0FAA/i5gGAA/GHeqlB/TM
---------------------------------------------------------------------~->

Send BUG REPORTS to bugs@xxxxxxxxxxxxx
Send SUGGESTIONS to suggest@xxxxxxxxxxxxx
-----------------------------------------
Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx 
(Web page: http://groups.yahoo.com/group/amiquote/messages/)
--------------------------------------------
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html 

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