Re: [OPE-L] The phenomenology of prices

From: Dave Zachariah (davez@KTH.SE)
Date: Mon Dec 31 2007 - 09:44:13 EST


I'm no expert on the subject. But I can't see any specific inherent
limitations of I/O modeling. One could describe a state transition from
time t to t+1, i.e. the update of the vectors and matrices, it may be
very hard but I think it is possible in principle.

What do you think are its limitations?

//Dave Z


on 2007-12-31 02:49 GERALD LEVY wrote:
> To my knowledge I/O models do not predict price variations, but if you
> > want a descriptive model of that then one could formalize it as discrete
> > time varying vectors and matrices on, say, a monthly basis.
>
>
>
> Hi Dave:
>
> Yes, I imagine that's possible if you have the data.  But, the question
> remains: is it possible to develop a genuinely 'dynamic I/O model'?
> The question I am trying to get at is: what are the inherent limitations
> of I/O modeling?
>
> In solidarity, Jerry


This archive was generated by hypermail 2.1.5 : Wed Jan 02 2008 - 00:00:07 EST