Let us suppose that we have a system a “free banking”
in which, unfortunately, fractional-reserve banking is not treated as
fraud. To what extent can banks use
fractional-reserve banking? What limits
does the market naturally place on it?
Firstly, for any bank to even begin
fractional-reserve banking, it must be reputable and trusted by individuals in
society. The bank must build up trust
among the people that it will be able to redeem their bank notes and bank
accounts at any time, on demand.
A second limit is the extent to which people use
banks. If individuals insisted on using
gold (or government paper), the opportunity for fractional-reserve is severely
reduced.
These limits are minor considerations, which mean
precious little once banks and banking have become established.
A much more powerful market restraint is the dreaded bank run. This occurs when clients of the bank – depositors
and/or note-holders – lose confidence in the bank, and begin to fear that the
bank does not have the money to redeem all its bank notes and deposit
slips. The clients begin to rush to the
bank to demand their money and, in the case of a fractional-reserve bank, there
is of course not enough money to redeem all outstanding notes and
deposits. Furthermore a run on one bank
can lead to runs on other banks, and a widespread loss of confidence can
quickly bring down the entire system and the game, for the bankers, is up.
Fear of a bank run will cause banks to be more
cautious and to not let the fraction of reserves they hold get too low. With a very low reserve ratio, even a mild
loss of confidence could cause a bank to collapse. A mild loss of confidence, typically caused
by a period of inflation due to bank credit expansion, will cause the banks to
increase their reserves. Hence, the
money supply will tend to contract and there will be a recession and deflation
of prices. This cycle of bank credit expansion,
causing inflation, leading to a loss of confidence and then a bank credit
contraction, is known as the “business cycle”.
But even the fear of bank runs still allows
fractional-reserve banking to be practiced to a considerable extent. There is, however, an even more powerful,
day-to-day constraint in the market.
This is the limited clientele of the bank. Any single bank’s fractional-reserve
practices are limited to the extent that their bank notes are used as a medium
of exchange.
We saw previously the example of a bank (let us call
it the Wright Bank), that had 500oz of gold in its vault, and had pyramided 5000oz
of bank notes on top of it, giving it a huge profit from loans to entrepreneurs
and borrowers. But the process does not
stop there. Suppose Smith has borrowed
1000oz of Wright bank notes for a business project. He does not hold on to the bank notes, but
spends them, causing a ripple of inflation from the point at which he injected
the new money into the economy. Suppose
with his 1000oz of bank notes, he paid them to Jones. Now, what is Jones going to do with the
money? If he is himself a client of the
Wright bank, then he now simply has 1000oz extra on deposit at the Wright
Bank. So long as confidence is retained,
all is well.
But what if Jones is not a client of the Wright
Bank? What if he prefers using gold or
is a client of a different bank? Then he
will go to the Wright Bank and demand redemption. Or, he will deposit his Wright bank note in
his own bank, the Brown Bank, and that bank would demand redemption from the
Wright Bank. This is not because he does
not trust the bank, or because he has lost confidence in it, but simply because
he prefers not to use the Wright Bank’s services. In any case, it is clear that it is sudden
death for the Wright Bank. It cannot
redeem the 1000oz of bank notes, because it only has 500oz of gold in the
vault. It is immediately bankrupt and
out of business.
Why would the Brown Bank want to redeem the Wright
bank note? Why should it do anything
else? The banks are competitors, not
allies. The Brown Bank would not want to
issue the Wright bank note to its customers, and it does not want notes in its
vaults: it wants gold, ready to redeem to customers on demand. Furthermore, if the Brown Bank is a
fractional reserve bank as well, gold can be added to its reserves and it can
pyramid bank credit on top of it. There
is no reason for banks not to demand redemption for each other’s notes.
Thus as soon as one bank’s note is deposited in another
bank, it will be redeemed. It is
therefore clear that a bank with a large clientele will have more opportunity
for fractional reserve banking, because it will take longer before his bank
notes reach somebody who is not a client of that bank.
Consider the extremes. In the absurd case of every bank having
exactly one client, no fractional reserve banking is possible, because as soon
as the one client spends his loan in the form of an un-backed bank note, that
bank note would be redeemed immediately by the bank of the person who accepted it.
On the other hand, if every person in a whole country
used the same bank, this limit would be removed entirely, and the single bank
could inflate the currency as it saw fit, subject only to the other market
limits, like the inevitable loss of confidence necessitating a contraction and,
in the worst case, a bank run. But even
with one bank for a whole nation, there is still the risk of redemptions from
overseas. This is especially likely if
the bank expands to any great extent, because it will cause prices to rise at
home and hence foreign goods will tend to be imported, causing the bank notes
to quickly go overseas.
One way to avoid this problem entirely, of course, is
for all the banks to simply agree not to redeem each other’s bank notes, so
that they can practice fractional reserve banking together, and all get away
with it for as long as they could before the inevitable loss of confidence
causing a recession. In other words,
they could form a cartel agreement.
As with any cartel on a free market, a banking cartel
would be inherently unstable and unable to survive very long, due to both
internal and external factors causing the cartel to break down. The external factor is competition from new
banks, and sound banks not interested in forming a cartel. As these competitors would instantly ask for
redemption from whichever cartel bank it receives a note from, gold would flow
from the cartel banks to the sound banks, which would cause a contraction of the
cartel banks. The non-cartel banks could
cause a deliberate loss of confidence in the inflated banks, just by
advertising their own non-inflationary practice and pointing out the
unsoundness of the cartel banks. The
cartel would soon break down and the member banks would be forced to either
become sound or go out of business. Only
with government grants of privileges excluding competition would such a cartel
be protected from external pressures.
There would also be internal pressures. Why, for example, would a sounder bank
continue in the cartel, effectively enabling less sound cartel members to gain
greater profits and a bigger market share?
They may get around this by agreeing to all abide by a fixed reserve
ratio, say 10%, but during inflationary booms there will be a temptation for
all the banks to expand faster and break the agreement, and during recessions,
many banks, fearing for their own survival, will contract faster.
So, in a “free banking” system, there are limits to
the extent to which banks can practice fractional reserve banking. These are: