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

Re: The peculiar System



PureBytes Links

Trading Reference Links

Dear Tomasz,
Well, it would be better to define "realistic" settings.
Then we could have a common measure, and this will give a meaning to 
the backtesting procedure.
So, what is your opinion ?
Best Regards
Dimitris Tsokakis
--- In amibroker@xxxx, "Tomasz Janeczko" <amibroker@xxxx> wrote:
> Dimitris,
> 
> I was not commenting your system. I think it is interesting and
> I appreciate that you want to share with the others.
> 
> The only thing I wanted to point out is
> that you should use more realistic back testing settings
> (this applies to ALL systems).
> 
> The reason why I commented this particular e-mail
> was that you were surprised why after reversing buy and sell
> rules the system performed also quite well.
> You said "It is not easy to explain this property."
> 
> My answer was: unrealistic settings could (at least partially)
> explain that.
> 
> Best regards,
> Tomasz Janeczko
> ===============
> AmiBroker - the comprehensive share manager.
> http://www.amibroker.com
> 
> 
> ----- Original Message ----- 
> From: "DIMITRIS TSOKAKIS" <TSOKAKIS@xxxx>
> To: <amibroker@xxxx>
> Sent: Saturday, October 13, 2001 9:34 PM
> Subject: [amibroker] Re: The peculiar System
> 
> 
> > Dear Tomasz,
> > Your comparison is not that fair and it is extremely simplified.
> > If you read the text you replied, you will see that I had
> > 32 trades from Jan 2000.(~450 trading days).
> > This means 1 trade per 14 days.
> > The neutral buy-sell system would be
> > 
> > buy= cum(1)%14==1;
> > sell=ref(buy,-13);
> > 
> > ie 32 trades equally distributed in the whole period, whithout
> > any other criteria applied.
> > For your information, this system with the same settings for
> > the same stock and the same period, gives Net % profit -24%.
> > Best Regards
> > Dimitris Tsokakis
> > --- In amibroker@xxxx, "Tomasz Janeczko" <amibroker@xxxx> wrote:
> > > Dimitris,
> > > 
> > > The reason why you get so good results is the one
> > > I mentioned already sometime ago:
> > > you set unrealistic entry/exit settings - buy on day's low
> > > and sell on day's high. 
> > > 
> > > To prove it: please modify your selling formula to:
> > > 
> > > sell = buy; /* sell on the same bar as you buy */
> > > 
> > > and your system will be winning no matter
> > > what is a buy formula with buying day low and selling day's 
high!
> > > 
> > > You can go further buy on low and sell on high everyday - 
> > > This "system" generates 45272.59 % during 281 days on DJIA 
stocks
> > > without ANY optimization.
> > > 
> > > Great ? Not really - it is just an effect of unrealistic 
settings.
> > > 
> > > Best regards,
> > > Tomasz Janeczko
> > > ===============
> > > AmiBroker - the comprehensive share manager.
> > > http://www.amibroker.com
> > > 
> > > ----- Original Message ----- 
> > > From: Dimitris Tsokakis 
> > > To: amibroker@xxxx 
> > > Sent: Saturday, October 13, 2001 1:30 PM
> > > Subject: [amibroker] The peculiar System
> > > 
> > > 
> > > I wrote yesterday that this system is strange.
> > > One behavior never met before is the following:
> > > I scan with the exposed buy/sell conditions
> > > Buy=Cross(C2,C1);Sell=Cross(C1,C2);
> > > The best % Net profit was +391% and all the 120 combinations
> > > from optimization were profitable.
> > > Then I inverse buy/sell with sell/buy, ie
> > > Sell=Cross(C2,C1);Buy=Cross(C1,C2);
> > > The best % net profit was 88.4% (!!) and the half of the 120 
> > combinations
> > > were profitable, if for a period of
> > > two years you were making mistakes all the time (32 trades).
> > > It is not easy to explain this property.
> > > Dimitris Tsokakis
> > > 
> > > 
> > > Yahoo! Groups Sponsor 
> > > 
> > > 
> > > 
> > > Your use of Yahoo! Groups is subject to the Yahoo! Terms of 
> > Service.
> > 
> > 
> > 
> > 
> > Your use of Yahoo! Groups is subject to 
http://docs.yahoo.com/info/terms/ 
> > 
> > 
> >