Flexible exchange rates and the nned for reserves

[paul davidson <pdavidson@xxxxxxx>]

At 07:55 PM 4/21/02 , Henry Liu wrote: 
It seems that there is a confusion about the term flexible exchange
rate.
A freely exchangable currency with a traget exchange rate requires more
reserves 
to hold the target. If there is no target exchange rate, no reserves are
needed, 
as Gunnar points out.

Only under some very unrealistic conditions, can a logical construct that
makes the above point be demonstrated. (but as explained below these
conditions cannot exist in the real world.)

In the real world tGunnar's point is not correct even if the government
(or central bank) does not attempt to influence the exchange rate in any
manner,. Nevertheless, if there is to be an organized and orderly free
market in foreign exchange -- there must be "market makers" in
order to assure a liquid foreign exchange market. Thus private bankers
will have to move in to "make" the market and each of those
bankers will have to keep reserves to maintain orderliness. The sum total
of all the privately held reserves will exceed that necessary for central
banks to maintain a fixed exchange rate.  And these are NEEDED
reserves, not  "perceived"  reserves, to assure a
liquid, orderly market over time.

Moreover if one or more of the private banker "market makers"
were to go bankrupt in an attempt to maintain orderliness in the face of
a bandwagon (herd behavior) movement, then there need not exist any set
of exchange rates that will clear all markets simultaneously (as Arrow
and Hahn have, in their book, demonstrated -- any bankruptcy jeopardizes
the existence of any general equilibrium).

The idea that there can exist a free market determined equilibrium
exchange rate over time in a changing world economy is a Friedmanic myth
-- that requires impossible logical conditions -- as FrIedman admitted
when he responded to my criticism in the JPE in 1972 when Freidman
wrote" A price may be flexible... yet be relatively stable, because
demand and supply are relatively stable overtime (e.g., this was the case
with the exchange rate of the Canadian dollar in the 1950s). Violent
instability of prices in terms of a specific money would greatly reduce
the usefulness of that money...."

It is nice to know that exchange rates can be perfectly flexible as long
as demand and supply are stable over time..... The question occurs what
should be done when the exchange rate becomes unstable -- and people rush
to a different liquidity store of value?
All of this is discussed in great detail in a chapter of my forthcoming
book (to be published this summer) FINANCIAL MARKETS, MONEY AND THE REAL
WORLD.

Paul