Re: the Fed: comment from weblog

On 28 Jul 2005 13:44:05 -0700, "JKroeger" <james@xxxxxxxxxxxxx> wrote:

>>> term of the loan. The only interest rate the Fed controls is the
>>> overnight lending rate between banks, otherwise known as the Fed funds
>>> rate.
>
>>Don't take my word for it. Here is how McConnell & Bruce explained it
>>in their popular intro textbook:
>
>>"The Fed actually sets neither the Federal funds rate nor the prime
>>rate; each is established by the interactiona of lenders and borowers.
>>But the Fed is the monopoly suppier of bank reserves. Wht it reduces
>>bank reserves, the Federal funds rate goes up; when it increases bank
>>reserves, the Federal funds rate goes down.
>
>"The Fed sets a target for the Fed funds rate, and steers the average
>rate to its target quite closely through its open market operations,
>typically within 10 basis point on the weekly average. That is the
>sense in which the Fed "controls" the Fed funds rate."
>
>>And since the amount of
>>reserves in the banking system helps determine the total supply of
>>money, changes in the supply of bank reserves affect the money supply
>>and interest rates in general."
>
>"This is the standard textbook theory which fails to describe how
>things actually work. It implies that the Fed has a target for the
>money supply, and steers the money supply to its target by manipulating
>banking system reserves. The fact is the Fed long ago gave up
>targeting the money supply because it bore no particular relation to
>its primary targets -- inflation rate and unemployment. It now targets
>the overnight lending rate, and leaves both the money supply and
>banking system reserves as residuals. "
>
>I do not see the implication that you have inferred.
>
Specifically what is the implication I presumably inferred?

>The Fed's targets are well known.

The only Fed target that I mentioned is the Fed funds rate target.
What other targets do you have in mind?

>>The Fed Funds rate is 'set' by buying & selling short-term debt
>>instruments.
>
>>> Longer term rates are a function of many things, none of which the Fed
>>> directly controls.
>
>>The Fed can easily drive long-term rates down by simply buying up
>>long-term debt. If there are long-term notes being sold on the market,
>>the Fed can always be the highest bidder if it wants to acquire the
>>debt; if it offers higher than the going market price, it will drive
>>the interest rate higher. Driving long-term rates down is a different
>>sort of challenge I've explained elsewhere; I'll go over it here if
>>asked.
>
>"The Fed is constrained in what it can do to shape the yield curve by
>its need to control the Fed funds rate, i.e. the short term rate. It
>cannot arbitrarily add long term securities to its portfolio without
>selling short term securities, thus keeping the supply of reserves in
>line with the market for Fed funds."
>
>Please explain the basis of your claim that the Fed "cannot
>arbitrarily add long term securities to its portfolio without selling
>short term securities."

If the Fed _arbitrarily_ decided to buy long-term securities, it would
have to sell an approximately equal amount of its short-term
securities in order not to create an excess of banking system reserves
which would otherwise drop the bid rate on Fed funds to near-zero.

>Surely you are not suggesting that the Fed
>would have to sell short term securities IN ORDER TO OBTAIN THE FUNDS
>with which to buy the long-term securities?

The Fed has no constraint on the amount of funds at its disposal.

>I can only guess that you
>must be suggesting that a Fed decision to reduce the SUPPLY of
>long-term securities would necessarily affect market DEMAND for
>short-term securities. Even so, I don't see how this could possibly
>make any difference.

The problem has little to do with the availability of long and short
term securities. It has everything to do with the effect of the Fed's
purchase or sale of those securities on the amount of reserves in the
banking system. The Fed buys or sells only to maintain a balance in
supply and demand for reserves in the Fed funds market at its target
rate or interest.
>
>If a drop in the supply of long-term debt were to increase demand for
>short-term debt, the Fed could simply increase the supply of short-term
>debt to maintain the price it wishes. If a drop in the supply of
>long-term debt were to decrease demand for short-term debt, the Fed
>could simply decrease the supply of short-term debt to maintain the
>price it wishes. The Fed's unlimited control of supply in any
>debt-instrument market, gives it unlimited control the price of those
>securities. Where's the limitation that you are seeing?

No. The Fed does not have unlimited control of supply in the debt
instrument market. It holds a finite supply of securities in its own
portfolio which it can sell in order to drain banking system reserves.
It has unlimited funds with which to buy securities in the debt
instrument market. However it only does so as needed to add reserves
to the banking system as required to maintain control of the Fed funds
rate.
>
>>> Misleading. The Fed cannot control the quantity of reserves it
>>> creates without giving up control of the price of reserves, i.e. the
>>> Fed funds rate. Theoretically the Fed has unlimited spending power.
>>> However it cannot use that power by arbitrarily injecting reserves
>>> into the banking system. If it did so, it would lose control of the
>>> Fed funds rate.
>
>>Well, the whole point of 'arbitrarily injecting reserves into the
>>banking system' is to CONTROL the FF rate when 'market forces' would
>>otherwise drive the FF rate above the FOMC's target.
>
>"The key word here is "arbitrarily", which means without concern for
>its effects on other variables. Reserves cannot be injected
>arbitrarily."
>
>Because...?
>
Because the Fed cannot inject Fed funds _arbitrarily_ into the banking
system without affecting the Fed funds rate. It can do one or the
other, but not both simultaneously.
>
>"They can be added or drained only as required to
>"control" the Fed funds rate -- assuming the Fed wants to maintain
>control of the Fed funds rate, which has been its primary monetary
>policy instrument for decades."
>
>What mutually exclusive conditions or limiting parameters are you
>alluding to here?
>
As just explained, the Fed can control either the supply of reserves
or the "price" of reserves (the overnight interbank lending rate) but
it cannot control both simultaneously. They are mutually exclusive
variables.
>
>>>Loanable funds are created by the banking system, not by the Fed.
>
>>When the Fed buys bonds from banks, it gives them money that was not
>>saved by any saver. The banks can lend that money because it is a part
>>of their excess reserves. I really don't understand how Hummel can
>>take that information and conclude that the Fed does not create
>>loanable funds.
>
>
>"This is same the erroneous theory of money seen in standard econ
>textbooks. It assumes that the Fed initiates a sequence involving its
>injecting reserves, which are then loaned by banks. The sequence is
>backwards. In truth, the Fed buys bonds from banks (or security
>dealers) to satisfy the demand for reserves that banks as a whole need
>to back their net lending. The reserves themselves are not loanable
>by definition. They are what the banks must hold to back their
>lending."
>
>Let me see if I'm understanding you correctly. A commercial bank
>makes some very profitable loans but finds that it will not meet
>reserve requirements after losing 15% of its depositors to a competitor
>that day. It decides to sell most of its Treasury holdings to cover
>the reserve requirement. Are you telling me that because the proceeds
>from the bond sale are going to become part of the bank's required
>reserves, there is no injection of the Fed's "keystroke money"
>into the economy?

The bank can sell its liquid assets like T-bills to raise the funds it
needs to meet the reserve requirements -- if it has enough T-bills.
>
>What if the sequence of events was different? What if the bank would
>have been able to meet reserve requirements in spite of the loss of
>depositors if it hadn't made those profitable loans earlier in the
>day? What if, as the end of the business day approached, the bank
>found out that it could make those profitable loans but that it
>didn't have enough excess reserves to lend after the loss of
>depositors. If it then sold its Treasury holdings in order to
>capitalize the new profitable loans, would its receipt of the Fed's
>"keystroke money" and its subsequent lending of that money
>represent an "increase in loanable funds" to you?

A bank with adequate capital can make a loan before it has the
required reserves, and borrow funds in the money market or from the
Fed to acquire the reserves. But note that the bank does not loan
reserves, rather it issues its own debt to make a loan.

>Since the Fed's
>money was not held in reserve but was actually injected into the
>economy?

Can't make sense of this.
>
>I'm wondering why you are claiming that banks would only sell bonds
>to the Fed (or to security dealers) in order to meet reserve and/or
>capital ratio requirements?

I didn't say that. Furthermore, selling bonds to the Fed has no
effect on a bank's capital ratio.

Normally the Fed on its own initiative buys bonds from security
dealers to add reserves to the banking system when it sees the Fed
funds rate rising. However if an individual bank needs to acquire
reserves in a hurry to provide sufficient clearing balances at the Fed
to cover its depositors withdrawals, it will either sell liquid assets
or borrow in the money market or from the Fed.

>Are you telling me that banks would never
>sell some of their bond holdings simply to make a profit?

Of course not. Where did I say that?

>If the Fed
>was offering a price for certain bonds that would guarantee at least
>some banks a healthy capital gain, they would never do so? I'll
>grant you that some loaned-up banks will sell bonds for cash for the
>express purpose of covering the Fed's reserve and/or capital
>requirements, but is it accurate to generalize as you apparently have
>from this example?

Where did I generalize in the way you have just claimed?
>
>If you wouldn't mind, I'd appreciate it if you were to reconcile
>your account & conclusions with the conceptual reality that whenever
>the Fed is a net buyer of bonds, pieces of paper are taken out of the
>economy and replaced with cash that banks will want to lend out as soon
>as they can. Prior to the moment when this keystroke cash is lent out
>by banks, it is part of the bank's loanable reserves, is it not?

To repeat, a bank does not need reserves to make a loan. Therefore it
is misleading to talk about a "bank's loanable reserves" which by
definition must be held, not loaned. Note that a bank's required
reserves relative to its transaction deposits are based on a two-week
average each. It can hold zero reserves on a given day if it has
enough reserves on average over a two-week time frame.
>
>>> >The Fed is the only determinant of money supply that matters.
>
>>> Quite wrong. The money supply is determined by the public's demand
>>> for bank loans and the willingness of banks to lend.
>
>>It would actually have been more precise for me to say that the Fed is
>>the only determinant of LOANABLE FUNDS and INTEREST RATES that matters.
>
>"As previously noted, the Fed only controls the short term interest
>rate. It does not determine the amount of loanable funds. The amount
>of loanable funds for any given bank is limited only by the capital
>ratio requirement, not the reserve ratio requirement. As long as
>banks have adequate capital, they can increase their lending.
>Banks can borrow funds in the money market or from the Fed to meet
>their reserve ratio requirement, but they can only grow their capital
>by retained earnings or by selling more bank shares."
>
>Please be kind enough to enlighten me as to the flaw in the following
>statement:
>
>A bank can lend all of its cash assets EXCEPT for when EITHER (1) its
>total lending would exceed its capital ratio requirement, OR (2) its
>total lending would exceed the level permitted by the Fed's reserve
>requirement. True? False?

First, a bank does not lend cash assets. It lends by creating a
liability for itself and acquires an equal asset, namely the loan.
The capital ratio requirement is a complex formula, but it basically
sets a lower limit on the ratio of its capital to its total assets.
Since a new bank loan increases a bank's assets without increasing its
capital, the effect is to lower its capital ratio. Ultimately the
bank is limited in further lending when that ratio drops to the limit.

If the bank has more than enough capital to meet the capital ratio
requirement (which is normally the case), it can issue a loan before
it has the required reserves. As explained earlier, the reserve
requirement is based on a two-week average of its reserves versus its
transaction deposits over the same period. The reserve requirement is
not a constraint on bank lending. A bank in good standing can always
borrow the reserves it needs to meet the reserve ratio requirement,
and can do so after the fact.
.

Relevant Pages

  • Where do the $940 billion in excess Bank Reserves on the Fed Reserve Balance sheet come from?
    ... The $940 billion in excess Reserves are substitute for Fed Funds ... billion or so, but bank Reserves at the ...
    (sci.econ)
  • Re: the Fed: comment from weblog
    ... >>> overnight lending rate between banks, otherwise known as the Fed funds ... >>But the Fed is the monopoly suppier of bank reserves. ... >>money, changes in the supply of bank reserves affect the money supply ... > its need to control the Fed funds rate, ...
    (sci.econ)
  • Re: the Fed: comment from weblog
    ... "The Fed actually sets neither the Federal funds rate nor the prime ... But the Fed is the monopoly suppier of bank reserves. ... The Fed cannot control the quantity of reserves it ... When the Fed buys bonds from banks, it gives them money that was not ...
    (sci.econ)
  • Re: Non-Banks versus Banks
    ... >>customer simply brings $1 billion cash that he saved to his local bank ... Where he gets the money makes no difference. ... aggregate banking system reserves will increase by $1 billion ... to changes in aggregate reserves over which the Fed has no control. ...
    (sci.econ)
  • Re: who will buy our debt?
    ... As I understand how this all works out is, if no one will buy the debt instruments of the United States, the Federal Reserve Bank will buy these instruments. ... China Losing Taste for Debt From the U.S. ... The declining Chinese appetite for United States debt, ... Christmas that China's $1.9 trillion foreign exchange reserves had actually ...
    (soc.retirement)