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

Re: Building a Model



PureBytes Links

Trading Reference Links


>  Ron Augustine wrote:             
> Ideally, your basic assumption that it is necessary to know when to
switch
> between a trending and a oscillating analysis platform would appear to be
> correct-- 
>              
> A deeper assumption that it is possible to develop a mechanical system
that
> will consistently produce profits may be fundamentally flawed...  :)


Hi Ron -

I was testing some of the new code from the list this evening and found
myself starting to get sucked into the same traps as I had in the past. I
saw your statement and it inspired me to add my thoughts.

SYSTEMS and METHODS:   I feel it boils down to one area -  No matter what
you use, there still needs to be a forecast of what "type" of market you
expect to occur in the future.  There are so many "types" of markets and of
varying degrees.

For example, let's say you have a range-chop type system. Well, there are
an INFINITE amount of degrees of chopping markets that vary both in time
and magnitude for their rhythms....both periodic and aperiodic. You have
long swings, short swings, long time durations short time durations, to
infinity. You really need a different set of parameters for each. A chop
system can get destroyed in a chop that it is not set up for.

Likewise, there are many degrees of trending markets too. Some are smooth,
some have tremendous sell-offs, some are part chop and trend, etc.

The point I'm making is that if you are able to predict what kind of
precise market you expect, then you can pick a particular system that will
kick ass in it. 

** But then, isn't predicting a future market "type" what we are trying to
do in the first place?  To me, systems almost seem secondary  vs:
forecasting the expected market "type". **

When a system is finally designed that can look ahead and forecast "market
type" and dynamically switch in appropriate parameters, then I will agree a
mechanical system can consistently produce profits. Until then, when
markets change, expect drawdowns to eat most systems alive. 

MY QUESTION TO EVERYONE:  Other than the over-used indicators that are
supposed to tell us what "type" of market we are PRESENTLY in, what methods
can you think of that will forecast what "type" of market to expect NEXT?

Over-Simplified Examples: 

Maybe after sharp climax we should look for a chop.....etc? -   How to
measure this ? 

After a long, drawn out chop, maybe next a trend?   - How to measure?   


Tom Cathey