PureBytes Links
Trading Reference Links
|
At 12:16 AM 6/1/01 +0200, you wrote:
>are you using weekly data?
>because you have less trades then i get?
I only had data for ARTL for 1991-2001, you went back to 1986.
Now we have this impressive result ... what can we do with it?
I tried it on the QQQ and got about 470 % for march 1999 to today - values
found by optimizing. I am sure this could be improved with stop loss
techniques. It seems to work on selected stocks for selected periods. How
can we make an exploration to find stock responsive to this system?
I used this for the QQQ:
LONG:
Periods:=7;
a:=1.2;
e1:=Mov(C,Periods,E);
e2:=Mov(e1,Periods,E);
e3:=Mov(e2,Periods,E);
e4:=Mov(e3,Periods,E);
e5:=Mov(e4,Periods,E);
e6:=Mov(e5,Periods,E);
c1:=-a*a*a;
c2:=3*a*a+3*a*a*a;
c3:=-6*a*a-3*a-3*a*a*a;
c4:=1+3*a+a*a*a+3*a*a;
res:=c1*e6+c2*e5+c3*e4+c4*e3;
C<res
SHORT:
Periods:=2;
a:=0.9;
e1:=Mov(P,Periods,E);
e2:=Mov(e1,Periods,E);
e3:=Mov(e2,Periods,E);
e4:=Mov(e3,Periods,E);
e5:=Mov(e4,Periods,E);
e6:=Mov(e5,Periods,E);
c1:=-a*a*a;
c2:=3*a*a+3*a*a*a;
c3:=-6*a*a-3*a-3*a*a*a;
c4:=1+3*a+a*a*a+3*a*a;
res:=c1*e6+c2*e5+c3*e4+c4*e3;
C>=res
Trades: 230/157/73
Didn't test out of sample.
Best,
Herman
>At 04:12 PM 5/31/01 -0400, you wrote:
>>At 07:09 PM 5/31/01 +0200, you wrote:
>>>herman,
>>>
>>>why does he use the "P" (data array identifier) in his system?
>>
>>No idea ... Tried C & P, no real difference.
>>
>>>ARTL, DAILY, date period: 6/23/1986-5/30/2001, trade delay = 0
>>>profit: 364.2258
>>>trades: 1142/833/309
>>
>>Trade delay = 0 no wonder I couldn't get his results! Its pretty hard to do
>>your analysis at the close and buy before the close ... I buy at next day
>>Open.
>>
>>Profit around $6000000000 (612138%). for the period 1991-2001. Trades:
>>165/128/37.
>>
>>Thank so much for pointing out the zero delay.
>>
>>Herman.
>>
>>>
>>>no errors of any kind reported.
>>>
>>>onno
>>>
>>>
>>>
>>>At 02:37 PM 5/30/01 -0400, you wrote:
>>>>I have been unable to duplicate the results shown on page 31, Figure 2, of
>>>>the June issue of TASC, for Aristotle (ARTL). The author uses the AT3
>>>>system and gives the results as -1.92 % with the plain T3 system and
>>>>237000% with the AT3 system (see figure 2). I requested help from the
>>>>author (Steve Burns) a few days ago but got no reply.
>>>>
>>>>Can anybody step me through the procedure?
>>>>This system appears to be a good starting point... unless the performance
>>>>with ARTL is just plain luck.
>>>>
>>>>Best,
>>>>Herman.
>>>>
>>>>Ps. I am using these formulas:
>>>>
>>>>Enter Long:
>>>>Periods:=5;
>>>>a:=RSquared(C,opt2);
>>>>e1:=Mov(P,Periods,E);
>>>>e2:=Mov(e1,Periods,E);
>>>>e3:=Mov(e2,Periods,E);
>>>>e4:=Mov(e3,Periods,E);
>>>>e5:=Mov(e4,Periods,E);
>>>>e6:=Mov(e5,Periods,E);
>>>>c1:=-a*a*a;
>>>>c2:=3*a*a+3*a*a*a;
>>>>c3:=-6*a*a-3*a-3*a*a*a;
>>>>c4:=1+3*a+a*a*a+3*a*a;
>>>>res:=c1*e6+c2*e5+c3*e4+c4*e3;
>>>>C<res
>>>>
>>>>Exit Long:
>>>>Periods:=5;
>>>>a:=RSquared(C,opt2);
>>>>e1:=Mov(P,Periods,E);
>>>>e2:=Mov(e1,Periods,E);
>>>>e3:=Mov(e2,Periods,E);
>>>>e4:=Mov(e3,Periods,E);
>>>>e5:=Mov(e4,Periods,E);
>>>>e6:=Mov(e5,Periods,E);
>>>>c1:=-a*a*a;
>>>>c2:=3*a*a+3*a*a*a;
>>>>c3:=-6*a*a-3*a-3*a*a*a;
>>>>c4:=1+3*a+a*a*a+3*a*a;
>>>>res:=c1*e6+c2*e5+c3*e4+c4*e3;
>>>>C>=res
>>>>
>>>>Enter Short:
>>>>Periods:=5;
>>>>a:=RSquared(C,opt2);
>>>>e1:=Mov(P,Periods,E);
>>>>e2:=Mov(e1,Periods,E);
>>>>e3:=Mov(e2,Periods,E);
>>>>e4:=Mov(e3,Periods,E);
>>>>e5:=Mov(e4,Periods,E);
>>>>e6:=Mov(e5,Periods,E);
>>>>c1:=-a*a*a;
>>>>c2:=3*a*a+3*a*a*a;
>>>>c3:=-6*a*a-3*a-3*a*a*a;
>>>>c4:=1+3*a+a*a*a+3*a*a;
>>>>res:=c1*e6+c2*e5+c3*e4+c4*e3;
>>>>C>=res
>>>>
>>>>Exit Short:
>>>>Periods:=5;
>>>>a:=RSquared(C,opt2);
>>>>e1:=Mov(P,Periods,E);
>>>>e2:=Mov(e1,Periods,E);
>>>>e3:=Mov(e2,Periods,E);
>>>>e4:=Mov(e3,Periods,E);
>>>>e5:=Mov(e4,Periods,E);
>>>>e6:=Mov(e5,Periods,E);
>>>>c1:=-a*a*a;
>>>>c2:=3*a*a+3*a*a*a;
>>>>c3:=-6*a*a-3*a-3*a*a*a;
>>>>c4:=1+3*a+a*a*a+3*a*a;
>>>>res:=c1*e6+c2*e5+c3*e4+c4*e3;
>>>>C<res
>>>>
>>>>opt2 = 0.1 - 15 in steps of 0.1
>>>>
>>>>Best,
>>>>Herman.
>>>>
>>>
>>>
>>
>
>
|