[OPE-L:3774] Re: predicting market prices

Paul Cockshot (wpc@cs.strath.ac.uk)
Wed, 4 Dec 1996 02:48:16 -0800 (PST)

[ show plain text ]

>I'm not sure what you mean here about "predicting market prices." I
>thought that your work attempts to show an empirical *correlation* between
>values, production prices, and market prices. This seems to me to be very
>different that *predicting* prices. That is, don't you take the labor
>values, production prices, and _ex post_ market prices as *data* rather
>than develop a predictive model for market prices?
>If you actually think that you can predict market prices, then I (along
>with many investors on Wall Street!) would like to know your "system".

you are right here of course, we are not predicting future prices, but
seeing how actual past prices for some 100 industries compare with
what different theories predict that they would have been.

If one had access to the data, detailed i/o tables, sufficiently early, then
I think that one could have a good stab at predicting future prices. The
problem would be to collect the data. Whilst it is technically quite
practical to collect the data with modern computer technology, this collection
involves a real cost in labour and equipment. Under capitalism there
is no way of meeting the social cost of gathering the information. I would
advocate that one of the first steps in socialising an economy should be
the establishment of the sort of procedures for gathering this data.
Paul Cockshott