- This topic is empty.
-
Longer term sounds good. My thoughts are given that the 2 hottest books out there are written by Gary Antonacci and Andreas Clenow, why don’t we start with these two and start testing them on a couple of markets. While my coding is not fantastic I’m willing to do some checking and do some of the stress testing. I have both of them and am willing to let others borrow them in the name of education.
Maurice Petterlin wrote:Longer term sounds good. My thoughts are given that the 2 hottest books out there are written by Gary Antonacci and Andreas Clenow, why don’t we start with these two and start testing them on a couple of markets. While my coding is not fantastic I’m willing to do some checking and do some of the stress testing. I have both of them and am willing to let others borrow them in the name of education.A start from Clenow’s strategy in “Stocks on the move”:
Code:_SECTION_BEGIN (“System Parameters”);
//=================================================================================
//Parameters
//=================================================================================
// calculate the exponential regression, annulised and adjusted slopes
period = 90;
natLog = ln(C);
expSlope = LinRegSlope(natLog,period);
annulisedSlope = 100*((exp(expSlope)^250) – 1);
r2 = Correlation(natLog,BarIndex(),period)^2;
adjSlope = annulisedSlope*r2;// gap filter code
gapPercent = Param(“Gap %”,15,1,200,0.01);
gapPeriod = Param(“Gap Up Lookback”,90,1,1000,1);
priceField1 = C;
priceField2 = Ref(C,-1);
ratio = ((priceField1-priceField2)/priceField2)*100;
gap = (GapUp() OR GapDown()) AND abs(ratio) >= gapPercent;
gapNumber = Sum(gap,gapPeriod);
gapFilter = gapNumber == 0;// moving averages
longMALength = Param(“Long MA”,100,50,200,10);
atrLength = Param(“ATR length”,20,5,30,1);
longMA = MA(C,longMALength);//=================================================================================
//Entry & Exit
//=================================================================================
OnLastTwoBarsOfDelistedSecurity = BarIndex() >= (LastValue(BarIndex()) -1) AND !IsNull(GetFnData(“DelistingDate”));cond1 = C > longMA;
cond2 = gapFilter AND indexUp AND HDBFilter AND optionalFilter;setup = cond1 AND cond2;
entry = Ref(setUp,-1) AND NOT OnLastTwoBarsOfDelistedSecurity;Maurice Petterlin wrote:Longer term sounds good. My thoughts are given that the 2 hottest books out there are written by Gary Antonacci and Andreas Clenow, why don’t we start with these two and start testing them on a couple of markets. While my coding is not fantastic I’m willing to do some checking and do some of the stress testing. I have both of them and am willing to let others borrow them in the name of education.In the name of education here is just the strategy/algorithm rules from Clenow’s book:
I have one idea a bit weird one
but why we do not put the targets and start from thereCAR
MDD
trade frequency
holding timeI imagine it will be more effective than taking a strategy and testing it
Said Bitar wrote:I have one idea a bit weird one but why we do not put the targets and start from there
CAR
MDD
trade frequency
holding timeI imagine it will be more effective than taking a strategy and testing it
Yeah i agree here. Not sure about everyone else?
I was thinking a longer term model 3 months+ for holding time, so trade frequency will be low i guess. But that is just me…
Trent Rothall wrote:Said Bitar wrote:I have one idea a bit weird one but why we do not put the targets and start from there
CAR
MDD
trade frequency
holding timeI imagine it will be more effective than taking a strategy and testing it
Yeah i agree here. Not sure about everyone else?
I was thinking a longer term model 3 months+ for holding time, so trade frequency will be low i guess. But that is just me…
happy to contribute to any strategy. im planning on trading just one MRV strategy (hopefully start around july) until the end of the year before adding something else into the mix once i have bedded down a routine. a longer term strategy might be a good addition me.
- You must be logged in to reply to this topic.