From: Philip Dunn (hyl0morph@YAHOO.CO.UK)
Date: Sat Oct 20 2007 - 11:57:43 EDT
The next Mohun and Veneziani example:
"2.3 Example 3: 'Proof' that the temporalist MELT is initially positive
and finite (Kliman & Freeman 2006, pp. 122-3) Kliman and Freeman are
emphatic that the temporalist MELT [tau] is not undefined, because it is
the ratio of total price to total value. Rearranging equation (I),
[tau](t + 1) = [tau](t)P(t + 1)/C(t) + [tau](t)L(t) (5)
This serves to define the MELT of one period in terms of the preceding
period's MELT. For this to be a definition, an independent definition of
[tau](0) must be given. Kliman and Freeman conspicuously fail to do
this. They have no explanation of why [tau](0) is independent of
[tau](--1)--if it is not, there is an infinite regress; if it is, then
there must be some explanation of why [tau](1) is not independent of
[tau](0). None is forthcoming, and hence the TSSI MELT is undefined...."
I do not think an independent *definition* of [tau](0) is necessary. It
is already defined as the ratio of total price to total value for time
t=0. Operationally, it is enough to give an initial estimate. This has
been discussed here before:
http://ricardo.ecn.wfu.edu/~cottrell/OPE/archive/9811/0029.html
If you want to *measure* [tau](0) simply make an estimate of [tau](-20)
and crunch the numbers.
Phil
PS
Eqn 5 is a bit clearer written like this:
[tau](t + 1) = [tau](t)P(t + 1) / { C(t) + [tau](t)L(t) } (5)
___________________________________________________________
All new Yahoo! Mail "The new Interface is stunning in its simplicity and ease of use." - PC Magazine
http://uk.docs.yahoo.com/nowyoucan.html
This archive was generated by hypermail 2.1.5 : Wed Oct 31 2007 - 00:00:19 EDT