Re: Estimating money multiplier

["Henry C.K. Liu" <hliu@xxxxxxxxxxxxxx>]

Money Multiplier
• The ratio of a money aggregate to base
money (M0)
• M1 multiplier = M1/Monetary Base
• Ranges from 3.1 to 1.7 in US

As any student of money and banking knows, up to now our monetary system
has been one in which the money multiplier--the ratio of total public
deposit and currency holdings to the monetary base (the outstanding
amount of Federal Reserve Notes and bank reserve credits at the
Fed)--depends on at least two variables. These are (1) the public's
desired currency-to-deposit ratio (c) and (2) the bank's desired
reserves-to-deposit ratio (r). The formula for the multiplier is m = (1
+ c)/(r + c), where the total money stock, M, is equal to mB, and B
stands for the monetary base. In this formula, B is the only thing that
the Fed controls with any degree of precision. The great virtue of a
monetary base rule is, therefore, that the Fed could not fail to abide
by such a rule except through outright negligence or caprice. In
contrast, with any other sort of rule (including a zero inflation rule),
the Fed could always plead unforeseen circumstances if it failed to keep
its promise.

A long-standing argument against a monetary base rule is that such a
rule would not allow the central bank to adjust the base in response to
unforeseeable changes in the currency ratio. Unpredicted changes in c
would then lead to undesired changes in the money stock, nominal
spending, and prices. The emergence of e-money strengthens the case for
a strict monetary base rule by, in effect, setting the stage for
removing the currency-ratio as a factor in the money-multiplier. The
multiplier would then be simply 1/r--the reciprocal of the banking
system reserve ratio. The challenge of monetary control would be
simplified accordingly: With one less variable to worry about, the Fed
would have one less reason to improvise.
From:
http://www.cato.org/moneyconf/14mc-5.html

See also:

http://research.stlouisfed.org/fred2/series/MULT/

Clifford Poirot wrote:
> I am curious if anyone knows (or knows where to find) the type of model
> central banks use to estimate money multipliers. I would also be interested
> if anyone has done any recent work on the usefulness of the money multiplier
> model, particularly from a Post-Keynesian perspective.
>
> In the interim, a few questions:
>
> 1. From a Post-Keynesian perspective, does the high powered money
> (multiplier)=Ms model have any utility whatsoever (I suspect I would get
> different answers from the hard core endogenous money theorists and from
> those who do not think money is entirely endogenous)?
>
> 2. Is it possible to actually estimate a money multiplier from time series
> data if you have uncertainty?
>
> 3. Would this estimate, have any utility in use in trying to hit money
> supply and interest rate targets?
>
> I have my own biases and suspicions on these questions but I am interested
> in what others think, and specifically, on any Post-Keynesian or other
> recent work on this.