- 43 -
shrinkage estimation methodology produces very small or
very large errors at the aggregate level. Shrinkage is
not accrued at the aggregate level.
Dr. LaRue believes that, notwithstanding a net figure of zero at
the aggregate level, the fact that retail dollars of shrinkage
have to be converted into cost dollars among the various LIFO
pools, which pools utilize diverse conversion processes (i.e.,
differing BLS indexes and cost complements), undermines the
significance of Dr. Seago's shrinkage accrual accuracy analysis.
Dr. LaRue acknowledges, however, that a finding of low shrinkage
estimation error at the aggregate level is not irrelevant.
Indeed, he believes that such a finding creates a presumption of
a “pretty good model”.
In addition, Dr. LaRue, although in agreement with
Dr. Seago's mathematics, questioned the significance of the high
correlation between sales and shrinkage derived from the 10-year
correlation analysis. Dr. LaRue conducted his own analysis of
Target data and concludes that the correlation between sales and
shrinkage disintegrates as the level of analysis moves from the
Target-wide level to the department and store levels.
G. Evaluation of Expert Testimony
1. Origin of Shrinkage Estimation Errors
An inquiry into the origin of shrinkage estimation errors
provides the proper perspective to evaluate Dr. LaRue's criticism
that the Divisions' shrinkage methods are inherently flawed
Page: Previous 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 NextLast modified: May 25, 2011