[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[EquisMetaStock Group] Re: Rules Management for Trading System



PureBytes Links

Trading Reference Links


Hi Eric,

I'm not entirely sure how to answer this, it is difficult since you
are combining different systems to generate the signals.  It could
potentially lead to problems, even if the same number of trades are
generated per year, since they can be vastly out of sync.  I think you
are fairly safe if you combine them the way we discussed before, but
best is to test the thing and see if you are happy with it in
practise.  It has a nice ring to it to use different long and short
systems - so I am not saying you should not do it this way, but just
verify the results you get from the tickers you are interested in by
eyeballing the data.

> p1 = 20% ROI then w1 = 1000
> p2 = 20% ROI then w2 = 1000
> p3 = 40% ROI then w3 = 2000 [which need to adjust manually]
> p4 = 20% ROI then w4 = 1000

Yes, exactly!  If you have enough data, the weights will eventually
react to the different ROIs and will adjust themselves, but the better
starting values you enter, the least data will be wasted on settling
down and the quicker you will get to the system's steady state.

Sometimes you can override the starting values also to ensure the
overall system makes use of a given subsystem.  Say there is a trend
following system that is slow to adjust and its weight becomes very
small, but you want it in at all times.  Then you can let its weight
not adjust or adjust slower than the rest.  We often have such a trend
following system as 'anchor' and either switch it on or off this way -
normally it has a very poor performance, but it still adds to the
overall system as it models the loooong term trend and the system
becomes based on deviations from this trend if it is included.

Regards
MG Ferreira
TsaTsa EOD Programmer and trading model builder
http://www.ferra4models.com
http://fun.ferra4models.com 


--- In equismetastock@xxxxxxxxxxxxxxx, chichungchoi <no_reply@xxxx> wrote:
> 
> Thank you for everyone :>
> 
> Hi MG Ferreira:
>   
> SHORT signal: RSI(9) - 70; [On average, there are 20 trades per year]
> LONG signal : C-Mov(C,25,S)[On average, there are 20 trades per year]
> 
> Even through the faster RSI(9) will normally dominate in term of 
> shorter periods, both triggered signals for LONG and SHORT are equal 
> in term of number of trades per year. Can it still be no problem to 
> combine them into one indicator? If trying to combine LONG and SHORT 
> signals into one, then should I consider the selection criteria based 
> on whether the number of periods for both signals or the same number 
> of average trades per year?  Do you have any suggestion?
> 
> On the other hands, in creating four weights, say w1 .. w4
> w1 = 1000
> w2 = 1000
> w3 = 1000
> w4 = 1000
> Do you have any suggestion on how to adjust the weight based on 
> indicator's reliability and performance? such as 
> Performance
> p1 = 20% ROI then w1 = 1000
> p2 = 20% ROI then w2 = 1000
> p3 = 40% ROI then w3 = 2000 [which need to adjust manually]
> p4 = 20% ROI then w4 = 1000
> 
> Will it be something like that? Any idea?
> 
> Thank you
> Eric





------------------------ Yahoo! Groups Sponsor --------------------~--> 
In low income neighborhoods, 84% do not own computers.
At Network for Good, help bridge the Digital Divide!
http://us.click.yahoo.com/EpW3eD/3MnJAA/cosFAA/BefplB/TM
--------------------------------------------------------------------~-> 

 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/equismetastock/

<*> To unsubscribe from this group, send an email to:
    equismetastock-unsubscribe@xxxxxxxxxxxxxxx

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/