[OPE-L:6633] [OPE-L:94] Price-Value Deviations

Gerald Levy (glevy@pratt.edu)
Wed, 9 Sep 1998 07:37:33 -0400 (EDT)

Date: Tue, 09 Sep 1997 12:50:08 +0300
From: Tsoulfidis Lefteris <lefteris@uom.gr>

Dear all,

I read the recent article by Steedman and Tomkins entitled: Measuring
the deviation of prices from values, in Cambridge Journal of Economics
1998, 22, pp. 379-385. In the article it is claimed that the deviation
between the vector of prices (Steedman and Tomkins mean prices of
production) and the vector of values depends on the the type of
numeraire that one chooses to use and they find that these deviations
may be much larger than those reported in various other studies (Ochoa,
Petrovic among others). Steedman and Tomkins further introduce a
statistic which, as they argue, does not depend on the type of
numeraire. The results that they report differ (sometimes substantially)
from those that one derives with the usual statistics of deviation,
i.e., MAD, MAWD, correlation coefficient, etc. which depend on the
selected numeraire. Can someone comment?