MOC & MR short systems

[danielbarbaro79]

Hi All , looking if i’m on the right track  with 2 of my systems i have crossed over from long to short with realtest
the direction i ended up going was the pretty good result all the time i.e. no index filter , minimum criterias ect . & filtered out  biotech  ,
so this worked for 1 each of my  MOC & MR but the stretched were very high to make the system work , like 1.2 or 1.5 ATR (did not work on all long systems)

results are pleasing the MR 5% CAR & 6DD , but negative correlation to long side, combined 35% car / 18% DD  
the MOC short  was pretty much the same car/dd as the long  version, only a tiny bit less exposure even with a giant stretch- unleveraged around 15% CAR 10% DD combined with the long version both at 100% exposure is around 30% CAR / 10% DD

both stress tested out of sample data and on small caps OK , paramater stability seems Good ,  im about ready to try a trial test with a mini account was just wondering if anyone else had taken a similar approach?

had some sucess with index filters and still looking there as well  , but the results were great some years , then long DD years in a row , compared to above where the results are flatter and more consistant year to year 

Hope this makes sence… really just looking to see if anyone else uses stretches this big & no index filters on short systems…
Thanks
Daniel

 

[JulianCohen]

I have a few strategies that don’t use Index filters. My stretch is around 0.6 though. I’m guessing your strategies are doing fewer trades with that stretch, which is fine. Probably a good win rate.

Something you could try..set an Index Filter and change the stretch if the Index Filter is on or off

Stretch: if(IndexFilt, 0.7, 1.2)

[Nick Radge]

My long/short day trade strategy doesn’t use an index filter. 

I do use one on the multi day strategy though

[danielbarbaro79]

Thanks Guys ,
Julian , the dual stretch made about a 25% improvement in both CAR and DD on the MR short , also optimizing the short (No of) positions with long and short combined added quite a bit to car and not DD(another 4 or 5 % car) . this was not apparent when just looking at the short system in isolation .

for my MOC , was using L + for entry , so the stretch was not as high as it 1st appeared . just checking now the differences using C & H for the entry

[JulianCohen]

Quote:

 the dual stretch made about a 25% improvement in both CAR and DD on the MR short , also optimizing the short (No of) positions with long and short combined added quite a bit to car and not DD(another 4 or 5 % car) . this was not apparent when just looking at the short system in isolation .
 

If you are backtesting InMEL with shorts then you might not get the fills in real time. I found after a while that it was better to trade shorts only InRUI as the probability of a fill was higher, therefore more chance to match the backtest

[danielbarbaro79]

just using RUI , will play about a bit more Tuesday.
the MOC works equally well with stretch set off L or C , not quite as well on high, thanks again for the input

[danielbarbaro79]

Ended up with index filter on / off
MOC short stretch .3 or .9
MR short stretch .9 or 1.2

[Author]

Posts