ALTERNATIVE MODELS OF BUFFER STOCK MONEY: AN EMPIRICAL INVESTIGATION
| DOI | http://doi.org/10.1111/j.1467-9485.1988.tb01027.x |
| Published date | 01 February 1988 |
| Author | V. A. MUSCATELLI |
| Date | 01 February 1988 |
Sconrrh
Journal
of
Political
Economy.
Vol.
35.
No
1,
Febmary
1988
0
1988
Scottish
Economlc Soclcty
ALTERNATIVE MODELS
OF
BUFFER
STOCK MONEY: AN EMPIRICAL
INVESTIGATION
V.
A.
MUSCATELL]*
University
of
Glasgow
The purpose of this paper is to examine some of the aspects
of
recent
theoretical models of buffer stock money, and to propose an alternative
framework which has certain desirable theoretical features. We test
the
different models on UK data for
A41
and show that some empirical support
exists for departing from the traditional approach to the modelling of buffer
stock money. Furthermore, some additional questions are raised about the
empirical modelling of the demand for money as a “buffer asset” which seem
to have received little attention
in
the literature.
We begin with a general introduction to the subject
of
“buffer stock
money”, proceeding
in
section two to the derivation
of
an alternative model
of
the demand for money. Empirical estimates of the alternative models for
the demand for
MI
in
the United Kingdom are reported in section three.
A
brief conclusion will then follow.
I
INTRODUC~ION
In recent years there have been a large number of theoretical and empirical
studies on the role of money as a “buffer stock asset”.’ The basic concept of
money as a “buffer stock” derives from the observation that money plays a
special role
in
individuals’ portfolios (see Goodhart,
1984).
Due
to the liquid
nature of money assets, the costs of adjusting money holdings are typically
less than the costs involved
in
changing holdings in real or illiquid financial
assets. In an uncertain environment, economic agents are likely to reallocate
their portfolios permanently only
if
they perceive permanent changes in those
variables affecting desired holdings of assets. Conversely, transitory changes
in
the economic environment are less
likely
to lead to such reallocations, with
money balances acting as a shock-absorber in such cases. For example,
consider the case where aggregate real income has suddenly risen. Unless
*
I
wish
to
thank John Foster, David Vines and an anonymous referee
for
their helpful
comments on an earlier draft
of
this paper.
I
am solely responsible
for
all remaining errors
and omissions.
‘Three useful surveys in this area are provided
by
Laidler
(1983).
Goodhart
(1984),
and
Cuthbertson and Taylor
(1986).
Date
of
receipt of final manuscript:
23
September
1987
1
2
V.
A.
MUSCATELLI
such a change is regarded as permanent, individuals are unlikely to reallocate
their portfolio
so
as to permanently augment their transactions balances. A
subsequent reversal of the rise in real income may result in potentially costly
short-term dealings in alternative illiquid financial assets. This approach
therefore leads to the assumption of forward-looking behaviour in the
demand for money.
In the United States studies were undertaken to examine the response of
the demand for money to anticipated and unanticipated shocks
in
the money
stock (see, for example Carr and Darby, 1981, MacKinnon and Milbourne,
1984). For example, in the Carr-Darby (1981) model forward-looking
behaviour takes the form of expectations about the future path of the money
stock. We will, however, concentrate on an alternative (but related)
approach, which differs from that proposed by Carr and Darby
in
its explicit
treatment of costs of adjustment, and the fact that economic agents do not
formulate their planned asset holdings on the basis of their expectations
regarding the path of an exogenous supply of money, but form a view
regarding the future path of the determinants of their asset demand functions
and, on the basis of this, plan their optimal holdings
of
money and other
assets over time (see for instance Cuthbertson 1985, Cuthbertson and Taylor,
1986). This approach is summarised in the following model, which forms the
basis of
a
number
of
empirical investigations of the demand for M1 (see for
instance Cuthbertson 1984, Cuthbertson and Taylor, 1987).
Assume that the desired, or “long-run” demand for money can be
described by the following equation:
M:
=
a0
+
alpt
+
CY~Y,
+
a3Rt
(1)
where
M*
represents the natural logarithm of the desired money stock,
R
the
interest rate on alternative assets, and
y
and
p
the natural logarithms of real
income and the price level respectively.2 Following our previous argument, if
there are costs of adjustment in reallocating the portfolio the representative
economic agent will attempt to find the optimal path for his actual money
balances,
MI,
over his time horizon. Assuming for simplicity that the
individual has an infinite time horizon, this choice may be characterised by
the minimisation of an intertemporal quadratic loss function,
C,
conditional
on information at time
t
-
1:
OD
c
=
~t-lz
a/ao(Mt+j
-
M:+,)*
+
al(Mt+j
-
Mt+j-1)2)
(2)
j=O
where
a.
and
a,
represent the weights attached to the costs of being away
from the desired long-run holdings of money, and the costs of adjusting
money holdings respectively, and
6
represents a subjective discount rate.
This type
of
optimisation
is
commonplace in other areas
of
economics,
including the modelling of a
firm’s
dynamic investment and labour demand
decision under uncertainty (see Sargent, 1979). Furthermore, we assume,
’The definitions
of
the data used
in
this
study are reported
in
the Data
Appendix.
To continue reading
Request your trial