Introduction

[JulianCohen]

Trent I created a long/short system from your template (thanks for that) and some of the examples.
However I have hit a snag. I want to rank the candidates by ROC. If I do that in Amibroker then I’d set Rank = 1000 – ROC for a long system and 1000 + ROC for a short system.

However this only has ranking in the main template. Can I put ranking into each system instead? If I do that, do I have to set a different rank in the main template?

[TrentRothall]

yes, remove the Setupscore from base template and add it in each strat.

The code you have will enter MaxPositions (numpos) in both strategies if you didn’t know. use
MaxPositions: NumPos – Combined(S.Positions ) if you want total positionsin both the = numpos

[TrentRothall]

i believe once you use the code

Using: base

In a strat it will inherit all those settings. so whatever is in base will be inherited into that strategy

[JulianCohen]

Like this ?

[JulianCohen]

Or should MaxSetups be

NumPos – Combined(S.Positions) ??

[TrentRothall]

umm not sure actually. you can use the ‘test log’ drop down to give you a good idea of how many setups and why positions were missed during a back test

[JulianCohen]

Thanks a lot Trent

[JulianCohen]

I noticed Marsten has a Biotech Universe Filter in his mean reversion systems. Discussing with Nick he said it was due to non FDA approval for a drug in testing having a massive impact on the stock at the next days open, which makes sense.

However, like many things that make sense, I tested on my MR and MOC systems and removing Biotech had a negligible impact on the bottom line. If anything it was adverse on the Swing system, which is where I would expect to see the most impact as you might well have been holding a stock doing well only to find it crashing whilst you are holding it.

I also tested on my Momentum system and again, minimal if adverse impact.

[SaidBitar]

I don’t know if it’s only me but there is difference with the results with amibroker

I think the issue is that in amibroker i am filtering based on price and volume (un adjusted) while in this one i am using the adjusted so all the results are different

Anyone noticed this?

[JulianCohen]

Said Bitar wrote:

I don’t know if it’s only me but there is difference with the results with amibroker

I think the issue is that in amibroker i am filtering based on price and volume (un adjusted) while in this one i am using the adjusted so all the results are different

Anyone noticed this?

Did you try switching Amibroker to adjusted to see if the trades match then?

[SaidBitar]

didn’t think of this
now i will do it

[SaidBitar]

yes after i set to adjusted the results are close

[JulianCohen]

Said Bitar wrote:

didn’t think of this
now i will do it

Always happy to help

[JulianCohen]

I’m just about to start scanning and working out how to create a DDE to go into SMART API.

I’ll let you know how I go.

[TrentRothall]

You can change the settings in RT see ‘adjustment’ in the guide Said. How are you going with the CSV Julian. You can print a string in RT i found out

[Author]

Posts