Books / Digital Text

X. Central Banking: Determining Total Reserves

The crucial question then is what determines the level of total bank reserves at any given time. There are several important determinants, which can be grouped into two classes: those controlled by actions of the public, or the market; and those controlled by the Central Bank.


The major action by the public determining total bank reserves is its demand for cash.1 We saw (in chapter IX and in Figures 9.1–9.7) how the public’s increased demand for cash will put contractionary pressure on a bank, while decreased desire for cash will add to its inflation of the money supply. Let us now repeat this for the aggregate of commercial banks. Let us assume that the public’s demand for cash in exchange for its demand deposits increases. Figure 10.1 shows a hypothetical banking system, and Figure 10.2 shows the immediate effect on it of an increase in the public’s demand for cash, that is, their redeeming some of its deposits for cash.


The hypothetical banking system is depicted as one with a 20 percent reserve ratio, fully loaned up. “Reserves” in the commercial banks’ asset column are of course exactly equal to “Demand deposits to banks” in the central Bank’s liabilities column, since they are one and the same thing. The asset side of the central Bank balance sheet is not being considered here; in our example, we simply assume that central Bank notes outstanding in the hands of the public is $15 billion. Total money supply in the country, then, is Demand deposits plus Central Bank notes, or

$50 billion + $15 billion = $65 billion

Now let us assume that the public wishes to draw down its demand deposits by $2 billion in order to obtain cash. In order to obtain cash, which we will assume is Central Bank notes, the banks must go to the Fed and draw down $2 billion worth of their checking accounts, or demand deposits, at the Fed. The initial impact of this action can be seen in Figure 10.2.


In short, depositors demand $2 billion in cash; the banks go to the Central Bank to buy the $2 billion; and the Central Bank, in exchange, prints $2 billion of new notes and gives them to the banks.

At the end of Step 1, then, the money supply remains the same, since demand deposits have gone down by $2 billion but Central Bank notes outstanding have increased by the same amount. The composition of the money supply has been changed but not yet the total. The money supply is still $65 billion, except that there is now $2 billion less of demand deposits and $2 billion more of Central Bank notes in the hands of the public.

But this is only the first step, because the crucial fact is that bank reserves have also gone down by $2 billion, by the same amount that Central Bank notes in the hands of the public have increased.

But since reserves have gone down, and the banks keep fully loaned up, this means that banks must contract their loans and demand deposits until the new total of deposits is again brought down to maintain the legal reserve ratio. As a result, bank loans and investments must contract by another $8 billion, so that the fall in reserves can be matched by a fivefold fall in total deposits. In short, the $2 billion drop in reserves must be matched by a total of $10 billion drop in demand deposits. At the end of the completed Step 2, therefore, the balance sheets of the banks and of the Central Bank look as follows (Figure 10.3).


The eventual result, then, of an increased demand for cash by the public is a drop in demand deposits of $10 billion, resulting from the drop of bank reserves of $2 billion. The total money supply has gone down by $8 billion. For demand deposits have fallen by $10 billion, and cash in the hands of the public has risen by $2 billion, making a net drop of $8 billion in the supply of money.

Thus, an increased demand for cash causes an equal drop in bank reserves, which in turn has a money multiplier effect in decreasing total demand deposits, and hence a slightly less intense effect in cutting the total amount of money.

If the public’s demand for cash drops, on the other hand, and it puts more of its cash in the banks, then the exact reverse happens. Suppose we begin with the situation in Figure 10.1, but now the public decides to take $2 billion out of the $15 billion of Central Bank notes in its possession and deposits them in exchange for checking accounts. In this case, demand deposits increase by $2 billion, and the banks take the ensuing extra cash and deposit it in the Central Bank, increasing their reserves there by $2 billion. The $2 billion of old Central Bank notes goes back into the coffers of the Central Bank, where they are burned, or otherwise retired or liquidated. This situation is shown in Figure 10.4.


In short, the immediate result of the public’s depositing $2 billion of cash in the banks is that, while the total money supply remains the same, only changing the composition between demand deposits and cash, total bank reserves rise by $2 billion.

Receiving the new reserves, the banks then expand credit, lending new demand deposits which they have created out of thin air. They pyramid deposits on top of the new reserves in accordance with the money multiplier, which in our stipulated case is 5:1. The final result is depicted in the balance sheets in Figure 10.5.


Thus, the public’s depositing $2 billion of cash in the banks increases reserves by the same amount; the increase in reserves enables the banks to pyramid $8 billion more of deposits by increasing loans and investments (IOUs) by $8 billion. Demand deposits have therefore increased by $10 billion from the reduction in the public’s holding of cash. The total money supply has increased by $8 billion since Central Bank notes outstanding have dropped by $2 billion.

In short, the public’s holding of cash is a factor of decrease of bank reserves. That is, if the public’s holding of cash increases, bank reserves immediately decrease by the same amount, whereas if the public’s holding of cash falls, bank reserves immediately increase by the same amount. The movement of bank reserves is equal and inverse to the movement in the public’s holding of cash. The more cash the public holds, the greater the anti-inflationary effect, and vice versa.

The public’s demand for cash can be affected by many factors. Loss of confidence in the banks will, of course, intensify the demand for cash, to the extent of breaking the banks by bank runs. Despite the prestige and resources of the Central Bank, bank runs have been a powerful weapon against bank credit expansion. Only in 1933, with the establishment of the Federal Deposit Insurance Corporation, was the government of the U.S. able to stop bank runs by putting the unlimited taxing and counterfeiting power of the federal government behind every bank deposit. Since 1933, the FDIC has “insured” every bank deposit (up to a high and ever-increasing maximum), and behind the FDIC—implicitly but powerfully—is the ability of the Federal Reserve to print money in unlimited amounts. The commercial banks, it is true, are now far “safer,” but that is a dubious blessing indeed; for the “safety” means that they have lost their major incentive not to inflate.

Over time, one powerful influence toward a falling demand for cash is the growth of clearing systems, and devices such as credit cards. People then need to carry less cash than before.2 on the other hand, the growth of the underground economy in recent years, in order to avoid income taxes and other forms of government regulation, has required an increase in strictly cash transactions, transactions which do not appear on the books of any government-regulated bank. In fact, it is now customary for economists to try to gauge the extent of illegal, underground transactions by estimating the increase in the proportion of cash transactions in recent years.

The major movement in the public’s demand for cash is seasonal. Traditionally, the public cashes in a substantial amount of demand deposits before Christmas in order to use cash for tips or presents. This has a deflationary seasonal effect on bank reserves. Then, in January, the cash pours back into the banks, and reserves rise once again. Generally, the Fed keeps watch on the public’s demand for cash and neutralizes it accordingly, in ways which will be explored below.


As in the case of the demand for cash in the form of Central Bank notes, an increase in the public’s demand for gold will be a factor of decrease in lowering bank reserves, and a fall in the demand for gold will have the opposite effect. Under the gold standard, with a Central Bank (as in the U.S. from 1913 to 1933), almost all of the gold will be deposited in the Central Bank by the various banks, with the banks getting increased reserves in return. An increase in the public’s demand for gold, then, will work very similarly to an increased demand for Central Bank notes. To obtain the gold, the public goes to the banks and draws down demand deposits, asking for gold in return. The banks must go to the Central Bank and buy the gold by drawing down their reserves.

The increase in the public’s demand for gold thus decreases bank reserves by the same amount, and will, over several months, exert a multiple deflationary effect over the amount of bank money in existence. Conversely, a decrease in the public’s demand for gold will add the same amount to bank reserves and exert a multiple inflationary effect, depending on the money multiplier.

Under the present fiat standard, there are no requirements that the Central Bank redeem in gold, or that gold outflows be checked in order to save the banking system. But to the extent that gold is still used by the public, the same impact on reserves still holds. Thus, suppose that gold flows in, say, from south Africa, either from outright purchase or as a result of an export surplus to that country. If the importers from south Africa deposit their gold in the banks, the result is an increase by the same amount in bank reserves as the banks deposit the gold at the Central Bank, which increases its gold assets by the same amount. The public’s demand for gold remains a factor of decrease of bank reserves. (or, conversely, the public’s increased deposit of gold at the banks, that is, lowered demand for gold, raises bank reserves by the same amount.)

So far, we have seen how the public, by its demand for gold or nowadays its demand for cash in the form of Central Bank notes, will help determine bank reserves by an equivalent factor of decrease. We must now turn to the major instruments by which the Central Bank itself helps determine the reserves of the banking system.


One method by which the Central Bank expands or contracts total bank reserves is a simple one: it increases or decreases its outstanding loans of reserves to various banks. In the mid-nineteenth century, the English financial writer Walter Bagehot decreed that the Central Bank must always stand ready to bail out banks in trouble, to serve as the “lender of last resort” in the country. Central Banks generally insist that borrowing from them is a “privilege,” not a right conferred upon commercial banks, and the Federal Reserve even maintains this about members of the Federal Reserve System. In practice, however, Central Banks try to serve as an ultimate “safety net” for banks, though they will not lend reserves indiscriminately; rather, they will enforce patterns of behavior upon borrowing banks.

In the united states, there are two forms of Federal Reserve loans to the banks: discounts and advances. Discounts, the major form of Fed loans to banks in the early days of the Federal Reserve System, are temporary purchases (rediscounts) by the Central Bank of IOUs or discounts owed to banks. These days, however, almost all of the loans are outright advances, made on the collateral of U.S. government securities. These loans are incurred by the banks in order to get out of difficulty, usually to supply reserves temporarily that had fallen below the required ratio. The loans are therefore made for short periods of time—a week or two—and banks will generally try to get out of debt to the Fed as soon as possible. For one thing, banks do not like to be in continuing, quasi-permanent debt to the Fed, and the Fed would discourage any such tendency by a commercial bank.

Figure 10.6 describes a case where the Central Bank has loaned $1 million of reserves to the Four Corners Bank, for a two-week period.


Thus, the Central Bank has loaned $1 million to the Four Corners Bank, by opening up an increase in the Four Corners checking account at the Central Bank. The Four Corners’ reserves have increased by $1 million, offset by a liability of $1 million due in two weeks to the Central Bank.

When the debt is due, then the opposite occurs. The Four Corners Bank pays its debt to the Central Bank by having its account drawn down by $1 million. Its reserves drop by that amount, and the IOU from the Four Corners Bank is canceled. Total reserves in the banking system, which had increased by $1 million when the loan was made, drop by $1 million two weeks later. Central Bank loans to banks are a factor of increase of bank reserves.

It might be thought that since the loan is very short-term, loans to banks can play no role in the bank’s inflationary process. But this would be as simplistic as holding that bank loans to customers can’t really increase the money supply for any length of time if their loans are very short-term.3 This doctrine forgets that if outstanding bank loans, short-term or no, increase permanently, then they serve to increase reserves over the long run and to spur an inflationary increase in the money supply. It is, admittedly, a little more difficult to increase the supply of outstanding loans permanently if they are short-term, but it is scarcely an insurmountable task.

Still, partly because of the factors mentioned above, outstanding loans to banks by the Federal Reserve are now a minor aspect of Central Bank operations in the United States. Another reason for the relatively minor importance of this factor has been the spectacular growth, in the last few decades, of the federal funds market. In the federal funds market, banks with temporary excess reserves at the Fed lend them literally overnight to banks in temporary difficulties. By far the greatest part of bank borrowing of reserves is now conducted in the federal funds market rather than at what is known as the discount window of the Federal Reserve.

Thus, during the 1920s, banks’ borrowed reserves from the Federal Reserve were at approximately 4 to 1 over borrowings from the federal fund market. But by the 1960s, the ratio of Federal Reserve to federal funds borrowing was 1 to 8 or 10. As J. Parker Willis summed up, “It may be said that in the 1920s Federal Funds were considered a supplement to discounting, but that in the 1960s discounting had become a supplement to trading in Federal Funds.”4

To get an idea of the relative importance of loans to banks, on January 6, 1982, the Federal Reserve Banks owned $1.5 billion of IOUs from banks; in contrast, they owned almost $128 billion of U.S. government securities (the major source of bank reserves). Over the previous 12 months, member banks borrowing from the Fed had increased by $335 million, whereas U.S. government securities owned by the Fed increased by almost $9 billion.

If the Fed wishes to encourage bank borrowings from itself, it will lower the rediscount rate or discount rate of interest it charges the banks for loans.5 if it wishes to discourage bank borrowings, it will raise the discount rate. Since lower discount rates stimulate bank borrowing and hence increase outstanding reserves, and higher discount rates do the reverse, the former is widely and properly regarded as a proinflationary, and the latter an anti-inflationary, device. Lower discount rates are inflationary and higher rates the reverse.

All this is true, but the financial press pays entirely too much attention to the highly publicized movements of the Fed’s (or other Central Banks’) discount rates. Indeed, the Fed uses changes in these rates as a psychological weapon rather than as a measure of much substantive importance.

Still, despite its relative unimportance, it should be pointed out that Federal Reserve rediscount rate policy has been basically inflationary since 1919. The older view was that the rediscount rate should be at a penalty rate, that is, that the rate should be so high that banks would clearly borrow only when in dire trouble and strive to repay very quickly. The older tradition was that the rediscount rate should be well above the prime rate to top customers of the banks. Thus, if the prime rate is 15 percent and the Fed discount rate is 25 percent, any bank borrowing from the Fed is a penalty rate and is done only in extremis. But if the prime rate is 15 percent and the Fed discount rate is 10 percent, then the banks have an incentive to borrow heavily from the Fed at 10 percent and use these reserves to pyramid loans to prime (and therefore relatively riskless) customers at 15 percent, reaping an assured differential profit. Yet, despite its unsoundness and inflationary nature, the Fed has kept its discount rate well below prime rates ever since 1919, in inflationary times as well as any other. Fortunately, the other factors mentioned above have kept the inflationary nature of member bank borrowing relatively insignificant.6


We come now to by far the most important method by which the Central Bank determines the total amount of bank reserves, and therefore the total supply of money. In the United States, the Fed by this method determines total bank reserves and thereby the total of bank demand deposits pyramiding by the money multiplier on top of those reserves. This vitally important method is open market operations.

Open market, in this context, does not refer to a freely competitive as opposed to a monopolistic market. It simply means that the Central Bank moves outside itself and into the market, where it buys or sells assets. The purchase of any asset is an open market purchase; the sale of any asset is an open market sale.

To see how this process works, let us assume that the Federal Reserve Bank of New York, for some unknown reason, decides to purchase an old desk of mine. Let us say that I agree to sell my desk to the Fed for $100.

How does the Fed pay for the desk? It writes a check on itself for the $100, and hands me the check in return for the desk, which it carts off to its own offices. Where does it get the money to pay the check? By this time, the answer should be evident: it creates the money out of thin air. It creates the $100 by writing out a check for that amount. The $100 is a new liability it creates upon itself out of nothing. This new liability, of course, is solidly grounded on the Fed’s unlimited power to engage in legalized counterfeiting, for if someone should demand cash for the $100 liability, the Fed would cheerfully print a new $100 bill and give it to the person redeeming the claim.

The Fed, then, has paid for my desk by writing a check on itself looking somewhat as follows:


Pay to the Order of Murray N. Rothbard     $100.00

Mr. Blank
Federal Reserve Bank of New York

There is only one thing I can do with this check. I cannot deposit or cash it at the Fed, because the Fed takes only deposit accounts of banks, not individuals. The only thing I can do is deposit it at a commercial bank. Suppose I deposit it in my account at Citibank. In that case, I now have an increase of $100 in my demand deposit account at Citibank; the bank, in turn, has a $100 check on the Fed. The bank greets the check with enthusiasm, for it now can rush down to the Fed and deposit the check, thereby obtaining an increase in its reserves at the Fed of $100.

Figure 10.7 shows what has happened as a result of the Fed’s purchase of my desk. The key monetary part of the transaction was not the desk, which goes to grace the increased furniture asset column of the Fed’s ledger, but that the Fed has written a check upon itself. I can use the check only by depositing it in a bank, and as soon as I do so, my own money supply in the form of demand deposits goes up by $100. More important, my bank now deposits the check on the Fed at that institution, and its total reserves also go up by $100. The money supply has gone up by $100, but the key point is that reserves have gone up by the same amount, so that the banking system will, over a few months, pyramid more loans and demand deposits on top of the new reserves, depending on the required reserve ratio and hence the money multiplier.


Note that bank reserves have increased by the same amount (in this case, $100) as the Fed’s open market purchase of the desk; open market purchases are a factor of increase of bank reserves, and in practice by far the most important such factor.

An open market sale has precisely the reverse effect. Suppose that the Fed decides to auction off some old furniture and I buy one of its desks for $100. Suppose too, that I pay for the sale with a check to the Fed on my bank, say, Citibank. In this case, as we see in Figure 10.8, my own money stock of demand deposits is decreased by $100, in return for which I receive a desk. More important, Citibank has to pay the Fed $100 as it presents the check; Citibank pays for it by seeing its reserve account at the Fed drawn down by $100.


Total money supply has initially gone down by $100. But the important thing is that total bank reserves have gone down by $100, which will force a contraction of that times the money multiplier of bank loans and deposits, and hence of the total money supply.

Therefore, if open market purchases of assets by the Fed are a factor of increase of reserves by the same amount, the other side of the coin is that open market sales of assets are a factor of decrease.

From the point of view of the money supply it doesn’t make any difference what asset the Fed buys; the only thing that matters is the Fed’s writing of a check, or someone writing the Fed a check. And, indeed, under the Monetary Control Act of 1980, the Fed now has unlimited power to buy any asset it wishes and up to any amount—whether it be corporate stocks, bonds, or foreign currency. But until now virtually the only asset the Fed has systematically bought and sold has been U.S. government securities. Every week, the System Manager (a vice president of the Federal Reserve Bank of New York) buys or sells U.S. government securities from or to a handful of top private dealers in government securities. The System Manager acts under the orders of the Fed’s Federal Open Market Committee, which meets every month to issue directives for the month. The Fed’s system Manager mostly buys, but also sells, an enormous amount, and every year the accumulated purchases of U.S. Treasury bills and bonds drive up bank reserves by the same amount, and thereby act to fix total reserves wherever the Fed wishes, and hence to determine the total money supply issued by the banks.

One reason for selecting government bonds as the major asset is that it is by far the biggest and most liquid capital market in the country. There is never any problem of illiquidity, or problem of making a purchase or sale in the government securities market.

How Fed open market purchases have been the driving force of monetary expansion may be quickly seen by noting that the Fed’s assets of U.S. government securities, totaling $128 billion in January 1982, was by far the bulk of its total assets. Moreover, this figure contrasts with $62 billion owned in 1970, and $27 billion owned in 1960. This is roughly a 17 percent (uncompounded) annual increase in U.S. government securities owned by the Fed over the past two decades. There is no need to worry about the ever-shifting definition of money, the ever-greater numbers of Ms. All that need be done to stop inflation in its tracks forever is to pass a law ordering the Fed never to buy any more assets, ever again. Repeatedly, governments have distracted attention from their own guilt for inflation, and scapegoated various groups and institutions on the market. Repeatedly, they have tried and failed to combat inflation by freezing wages and prices, equivalent to holding down the mercury column of a thermometer by brute force in order to cure a fever. But all that need be done is one freeze that governments have never agreed to: freezing the Central Bank. Better to abolish central banking altogether, but if that cannot be accomplished, then, as a transitional step, the Central Bank should be frozen, and prevented from making further loans or especially open market purchases. Period.

Let us see how a government bond purchase by the Fed on the open market increases reserves by the same amount. Suppose that the Fed’s System Manager buys $1,000,000 of government bonds from private bond dealers. (Note that these are not newly issued bonds, but old bonds previously issued by the Treasury, and purchased by individuals, corporations, or financial institutions. There is a flourishing market for old government securities.) in Figure 10.9, we show the System Manager’s purchase of $1,000,000 in government bonds from the securities dealer firm of Jones & Co. The Fed pays for the bonds by writing a check for $1,000,000 upon itself. Its assets increase by $1,000,000, balanced by its liabilities of newly-created deposit money consisting of a check upon itself. Jones & Co. has only one option: to deposit the check in a commercial bank. If it deposits the check at Citibank, it now has an increase of its own money supply of $1,000,000. Citibank then takes the check to the Fed, deposits it there, and in turn acquires a new reserve of $1,000,000, upon which the banking system pyramids reserves in accordance with the money multiplier.


Thus, a Fed purchase of a $1,000,000 bond from a private bond dealer has resulted in an increase of total bank reserves of $1,000,000, upon which the banks can pyramid loans and demand deposits.

If the Fed should buy bonds from commercial banks directly, the increase in total reserves will be the same. Thus, suppose, as in Figure 10.10, the Fed buys a $1,000,000 government bond from Citibank. In that case, the results for both are as shown in Figure 10.10.

Here when the Fed purchases a bond directly from a bank, there is no initial increase in demand deposits, or in total bank assets or liabilities. But the key point is that Citibank’s reserves have, once again, increased by the $1,000,000 of the Fed’s open market purchase, and the banking system can readily pyramid a multiple amount of loans and deposits on top of the new reserves.


Thus, the factors of increase of total bank reserves determined by Federal Reserve (that is, Central Bank) policy, are: open market purchases and loans to banks, of which the former are far more important. The public, by increasing its demands for cash (and for gold under the gold standard) can reduce bank reserves by the same amount.

  • 1. In this chapter, we will assume that cash is the notes of the Central Bank.
  • 2. But note that our previous concept of “cash balances” includes not only cash but also demand deposits and any other form of money, whereas now we are dealing with the public’s demand for cash per se as against deposits or other forms of money.
  • 3. This was one of the tenets of the “banking school” of monetary thought, prominent in the nineteenth century and still held in some quarters.
  • 4. J. Parker Willis, The Federal Funds Market (Boston: Federal Reserve Bank of Boston, 1970), p. 62.
  • 5. Interest rates on Fed loans to banks are still called “discount rates” despite the fact that virtually all of them are now outright loans rather than discounts.
  • 6. On the penalty rate question, see Benjamin M. Anderson, Economics and the Public Welfare, 2nd ed. (Indianapolis: Liberty Press, 1979), pp. 72, 153–54. Also see Seymour E. Harris, Twenty Years of Federal Reserve Policy (Cambridge, Mass.: Harvard University Press, 1933), vol. 1, pp. 3–10, 39–48.