An analysis of how a deposit or withdrawal of currency to or
from a US bank affects the M1 money supply provides an
instructive example of how our monetary system works. On average
depositors withdraw more currency than they deposit, so let's consider the
withdrawal by an individual depositor. Assume the amount is $100 to
keep the arithmetic simple. The bank debits his account, thereby
decreasing aggregate transaction deposits by the same amount the currency in
circulation increases. That leaves M1 unchanged -- for the moment.
Since vault cash is part of a bank's reserves, the bank will have
lost $100 in reserves. If the bank had no excess reserves before the
withdrawal, it will need to acquire $90 in reserves to cover its remaining
transaction deposits. In seeking to borrow that amount in the Fed funds market,
its bid will cause the overnight interest rate on Fed funds to
increase. The Fed will notice the increase
and inject $100 into the banking system to counter the upward
pressure on the interest rate. Of course the actual scenario is not as neat and
precise as this, but we are only seeking to understand the principle.
Why $100 instead of $90? Because the Fed creates a new transaction deposit
when in buys securities to inject reserves. Whichever bank ends up with
the $100 deposit needs only $10 of new reserves to cover it. Thus it can
lend the other $90 in the Fed funds market, which is exactly what the original
bank needs. The end result is that aggregate transaction deposits are
restored to their prior level.
Other things equal, M1 increases due to withdrawal of currency from
banks when the Fed acts to maintain control of the short-term interest rate, its
normal monetary policy response. The same result applies for a deposit of
currency, but with the Fed's response reversed.
William