Rank: Advanced Member
Groups: Registered, Registered Users, Subscribers, Unverified Users
Joined: 10/28/2004(UTC)
Posts: 3,111
Location: Perth, Western Australia
Was thanked: 16 time(s) in 16 post(s)
|
Ok...
What you are attempting is data mining. I do not recommend it.... but here we go anyway!
The most simple way is to just copy the indicator code into the Buy and Sell criteria sections of the EST:
{Buy}
shortperiods:=Opt1;
longperiods:=Opt2;
signal:=Opt3;
myMACD:=Mov(Mov(C,shortperiods,E)-Mov(C,longperiods,E) ,signal,E)
Cross(myMACD,0)
{Sell}
shortperiods:=Opt4;
longperiods:=Opt5;
signal:=Opt6;
myMACD:=Mov(Mov(C,shortperiods,E)-Mov(C,longperiods,E) ,signal,E)
Cross(0,myMACD)
NOTE:
There are 6 optimising variables here. If you let each opt have, say, 10 values then that is one milion computations, running over 2000 stocks, thats 2 BILLION (2,000,000,000) computations that have to be made!
If you let the 4 MA values have say 20 values and the 2 signal MAs have 10 values then you are looking at 16 million computations, over 2000 stocks is 32 BILLION (32,000,000,000) computations!
Either way, its going to take some time, and burn some CPU's.
Secondly,
If you really think that MACD crossover system would make you a profit, that everyone would already be using it? I reckon it has been tested before by a thousand people. Other than a learning experience, I do not recommend fitting the values to match the curve.
$0.02
Hope this helps.
wabbit :D
|
