|
PureBytes Links
Trading Reference Links
|
By the way, here is an improved formula, to avoid decimal levels like
16.5 etc.
The range is somehow extended but, for some stocks it is useful.
[It permits D1, D2 to move in the 70% of the min/max distance]
Indeed, the optimal solution has the last short back in Oct 2001.
This short is profitable now ~22%.
In indicator builder, plot RRR, D1 and D2 to see the reason.
The last year RRR oscillates mainly from 5 to 15.
For the period Jan 2001 till now for example, [12,11,16] is cute and
you may consider it as a more "local" solution.
The last Short is very nice...
[Do not forget to replace default optimization values with the
results of optimization research.]
/*Smoothed D-ratio optimized trading system*/
R=500*(H-L)/(H+L);
Z=Optimize("Z",14,10,25,1);
RRR=DEMA(R,Z);
Rmin=LastValue(Lowest(R));RminF=floor(LastValue(Lowest(R)))+1;
Rmax=LastValue(Highest(R));RmaxC=ceil(LastValue(Highest(R)))-1;
RRRmin=LastValue(Lowest(RRR));RRRminF=floor(LastValue(Lowest(RRR)))+1;
RRRmax=LastValue(Highest(RRR));RRRmaxC=ceil(LastValue(Highest(RRR)))-
1;
Filter=1;
AddColumn(R,"D-ratio");
AddColumn(RRR,"DEMA D-ratio");
AddColumn(Rmax,"HIGHEST D-ratio");
AddColumn(Rmin,"LOWEST D-ratio");
AddColumn(RRRmax,"HIGHEST DEMA D-ratio");
AddColumn(RRRmin,"LOWEST DEMA D-ratio");
AddColumn(RRRmaxC,"HIGHEST DEMA D-ratio C");
AddColumn(RRRminF,"LOWEST DEMA D-ratio F");
D1=Optimize("D1",15,RRRminF,0.7*RRRmaxC,1);
D2=Optimize("D2",16,floor(1.3*RRRminF),RRRmaxC,1);
F1=RRR>=D2;F2=RRR<=D1;
Sell=F2;Buy=F1;
Buy=ExRem(Buy,Sell);Sell=ExRem(Sell,Buy);
Short=Sell;Cover=Buy;
Short=ExRem(Short,Cover);Cover=ExRem(Cover,Short);
Plot(RRR,"",1,1);Plot(D1,"",4,1);Plot(D2,"",5,1);
For many combinations the system prefers D1, D2 with similar or
identical values.
Dimitris Tsokakis
PS. To remind, once again, previous performance does not guarantee
future performance...
--- In amibroker@xxxx, "wingnut_1944" <dingo@xxxx> wrote:
> I was trying to be very thorough in the information I provided and
> only forgot the symbol.. Not bad for me!
>
> It is the always mysterious QQQ.
>
> Sorry about that!
>
> --- In amibroker@xxxx, "dtsokakis" <TSOKAKIS@xxxx> wrote:
> > You forget to mention the stock under optimization.
> > In many examples we have no trades for a long period, so, there
is
> > not something wrong at the first sight.
> > If for Z, D1, D2 you have had some nice trades, highly profitable
> > perhaps, the system "avoids" to change these levels and cancels
> some
> > periods.
> > For TTM [trade the market] systems I have met sometimes
> long "silent"
> > periods. If they are in the beginning of the test period, it
means
> > that the system fits better to the recent period. If the silent
> > period is the recent one, it is not a good sign. May be this
system
> > is not operating any more [it is somehow "out of fashion"]
> > I have numerous examples for both logics.
> > I prefer systems with "normal" distribution of profitable trades.
> > I avoid systems with two or three excellent trades the first half
> of
> > 2000, with a result +1000% and then many losers wich give a final
> > +300%.
> > Dimitris Tsokakis
> > --- In amibroker@xxxx, "wingnut_1944" <dingo@xxxx> wrote:
> > > Dimitris,
> > >
> > > I pasted the code listed below into the Automatic Analysis
window
> > and
> > > ran an optimization from Jan 1 2000 thru July 2 2002 with the
> > > following optimization results:
> > >
> > > Z = 15
> > > D1 = 14
> > > D2 = 16.5
> > >
> > > But there are no trades after Oct 4, 2001!
> > >
> > > Obviously I'm doing something wrong.
> > >
> > > My Settings are:
> > >
> > > Positions = long and short
> > > commissions = 20 points ( I am assuming that this is $20)
> > > Buy/sell/short/cover = open with delay of 1
> > > Max Stop Loss of 10% with exit at stop
> > > Profit target and trailing stop are disabled.
> > >
> > > What am I doing wrong?
> > >
> > >
> > > R=500*(H-L)/(H+L);
> > > Z=Optimize("Z",10,5,20,5);
> > > RRR=DEMA(R,Z);
> > > Rmin=LastValue(Lowest(R));RminF=floor(LastValue(Lowest(R)))+1;
> > > Rmax=LastValue(Highest(R));RmaxC=ceil(LastValue(Highest(R)))-1;
> > > RRRmin=LastValue(Lowest(RRR));RRRminF=floor(LastValue(Lowest
> (RRR)))
> > +1;
> > > RRRmax=LastValue(Highest(RRR));RRRmaxC=ceil(LastValue(Highest
> > (RRR)))-
> > > 1;
> > > D1=Optimize("D1",14,RRRminF,RRRminF+0.5*(RRRmaxC-RRRminF),1);
> > > D2=Optimize("D2",16.5,RRRminF+0.5*(RRRmaxC-RRRminF),RRRmaxC,1);
> > > F1=RRR>=D2;F2=RRR<=D1;
> > > Sell=F2;Buy=F1;
> > > Buy=ExRem(Buy,Sell);Sell=ExRem(Sell,Buy);
> > > Short=Sell;Cover=Buy;
> > > Short=ExRem(Short,Cover);Cover=ExRem(Cover,Short);
> > > Plot(R,"D-ratio",1,1);Plot(RRR,"DEMA D-ratio",7,8);
> > > Plot(D2,"Buy Level",5,1);Plot(D1,"Sell Level",4,1);
> > >
> > >
> > > --- In amibroker@xxxx, "dtsokakis" <TSOKAKIS@xxxx> wrote:
> > > > Sometimes it is good to estimate the efficiency of your tool.
> > > > Read
http://groups.yahoo.com/group/Trading_Systems/message/588
> to
> > > see
> > > > better.
> > > > Dimitris Tsokakis
|