My own Weekend Trend System.

[AlexStylianou]

Hi all, 

Doing my own US Weekend system (Don’t have Australian Data) hoping for some advice on ranking 20 positions to avoid selection bias. Returns are not bad but ideally I want to bump Mar up to 0.8. Not sure how much the ranking can improve Mar and Drawdown though. Below is the results when doing the 1/Close as this has been the most profitable choice thus far.  

1/Close;  Return 18.10% (I like) Drawdown 36.10% (Not good) Mar 0.5. Dates to achieve these results 01.01.2000 – 08.08.2022
The stats not as good neglecting the start of the 21st century.  01.01.2012 – 08.08.2022 14.45%. Drawdown 30.93%

Other ways the system has been ranked. 
1/Open;
1000 + (ATR(100)/C);
HHV(C,50)/C;
ROC(C,100);

Will include my settings as I have made basic errors in the past.

With Nicks help I am not focusing on one index to catch some of those smaller stocks. 

Advice on ranking to have a nice weekend trend system or anything else relevant will be appreciated. 
-Alex. 

 

[JulianCohen]

I’m ranking with 1/ATR(50) on a universe of RUA and price filter of C > 1

Actually I am using a combination of 1/ATR(50) and 1/C…. I have an MAR of 0.66  

(multiple * 1/ATR(50) + (multiple * 1/C) 

Honestly I’m not sure that too much higher in MAR is robustly achievable…

Edit….I just found an error in my code

I have a MAR of 0.86 so it is possible…give that combination a go and see how it works for you.

 

[AlexStylianou]

Thanks Julian changing price to 1 instead of 5 helped but the 1/ATR(50) was slightly less. If you have not I think you should have a go at the 1/Close; and see how it works for you.

Below with $1 instead of $5

2000 – now backtest is 20 % + Drawdown 38.62%.

2012- now backtest is 13.42% Drawdown -36.62%

Really not sure about the drawdowns. Will do some more research.

[JulianCohen]

Hi Alex

I am using 1/Close….but I am also using 1/ATR(50)

I use them in combination with each other. 

[AlexStylianou]

I see my bad Julian I will have a go.

[Author]

Posts