|
PureBytes Links
Trading Reference Links
|
Sounds like a million dollar question. The model we use does not
try to predict the probability of an up or down, it signals a buy
or sell when a type of trix (cannot give more details than this)
indicator crosses a type of signal line. As long as it is above
this signal, you are 'long' and short on the other side. This is
more the way a trading model works than predicting the next day
probability. If you are really looking for such a probability
model I suggest you look at neural nets. They are good at predicting
in such a short term and will change position much more often than
the trix model.
To get a feel for it, use MS to plot a TRIX indicator and then take
a moving average of this TRIX. Whenever they cross, you buy or sell.
If you like what you see, you can play around with it some more or
with a similar but different system (such as say RSI) just to get a
feel for its performance. If this is not what you have in mind, I
guess you have to keep on searching.
We optimise the parameters, such as length of TRIX, length of moving
average, type of moving average and so on. This is a whole topic on
its own, but just play around a bit with these parameters and
hopefully you'll make some progress.
Regards
MG Ferreira
TsaTsa EOD Programmer and trading model builder
http://tsatsaeod.ferra4models.com
http://www.ferra4models.com
--- In equismetastock@xxxxxxxxxxxxxxx, "rorster36" <rorster36@xxxx> wrote:
>
>
>
>
> Thanks for your input on this perhaps you could give me a little more
> insight into setting up this Trix type model ?
>
> Basically all i want to know is if the dow will be up or down on the
> next day ? as this is a 50/50 chance anyway anything that can nudge
> the odds in my favour should help greatly .
>
>
> --- In equismetastock@xxxxxxxxxxxxxxx, "MG Ferreira" <quant@xxxx>
> wrote:
> >
> > There is a difference between predicting and generating trading
> > signals.... Trading using EOD is a bit tough on the Dow. We
> > toyed around with a number of EOD systems some time ago, but found
> > much better results for it specifically if we used a weekly or even
> > monthly approach, for two reasons - it has a strong trend over time
> > and lots of noise, making EOD difficult. Hope it helps. For what
> > it is worth, currently our best EOD model on the Dow is a Trix type
> > model, where a Trix crosses a signal line. Second best is a T3
> model,
> > where we calculate the change in a type of T3 model. Using weekly
> > data, we got good results from a type of moving average crossover
> > system, and using monthly we use a type of Coppoch model.
> >
> > Regards
> > MG Ferreira
> > TsaTsa EOD Programmer and trading model builder
> > http://tsatsaeod.ferra4models.com
> > http://www.ferra4models.com
> >
> >
> > --- In equismetastock@xxxxxxxxxxxxxxx, "rorster36" <rorster36@xxxx>
> wrote:
> > >
> > >
> > > Ive been playing with MS8 for some time now and am simply trying
> to
> > > get a decent system for the dow jones industrial average .Im not
> > > asking for the world just some way of predicting if the dow will
> be
> > > up or down the next day and was just wondering if one of you guys
> > > could help by pointing me in the right direction .
> > >
> > > Many thanks
------------------------ 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/
|