A Plan for Monetary Reform

The purpose of this monograph is to explore the characteristics and feasibility of a monetary system in which banks must fully back their transaction deposits with reserves of the monetary base. We will refer to this as a fully-backed system. Many writers have proposed such a system, but few if any have shown how it would work. This writer believes a fully-backed system would be an improvement on fractional reserve banking. The reader is invited to judge for himself.

The intent here is not to condemn the fractional reserve system. On the whole it works well when bank loans are made in support of productive enterprise or consumer finance. Such loans expand the money supply in line with the needs of the economy. However the fractional reserve system puts enormous leverage in hands of the larger banks, and that power is at times misused. Far too much lending now goes to support purely speculative activity in the financial markets. That distorts the markets, inflates asset prices, increases the fragility of the financial system, and serves no useful purpose in the real economy. If bank regulators took a hard line in proscribing such lending, the criticism of fractional reserve banking would be largely undeserved.

A few writers claim that fractional reserve banking is legalized counterfeiting. This absurd view arises from a misunderstanding of money itself. It is a carry-over from the days when the only “real money” was thought to be valuable tokens like gold coins, or notes convertible on demand to gold coins. In truth the intrinsic value of the tokens is irrelevant. Their exchange value is what matters, and that is determined by the policies and actions of the central bank. The State has the authority to set the fractional reserve requirement on bank deposits to any value it chooses. It is worth noting that several industrial nations today impose no reserve requirement on their banks.

Our focus will be mainly on the monetary system of the U.S. Even if it were obvious to most that a system of fully-backed deposits would be superior to the fractional reserve system, the chance of its being adopted is very low. The banking industry wields enormous political power, and that would be the main impediment to enacting legislation for major reform. A fully-backed system would change the role of banks as we know them in a very significant way. It would also rearrange the landscape in the financial industry, and greatly reduce the financial and political power of bank officials.

We will first review the basic characteristics of money and the banking system. Then we will examine the fractional reserve system of the U.S. Finally we will investigate the properties of a fully-backed system, identify key issues that must be addressed, propose a plan, and attempt to assess its operation in practice.

Money can be defined as whatever is widely accepted as a medium of exchange. Of course to be accepted, it must be seen as a store of value. These basic properties, however, do not explain how something gains status as money and how it is to be measured. Keynes held that the primary concept in the theory of money is the unit of account. Throughout history, States have established what is to serve as legal tender. They have done so by (1) giving a name to unit of account; (2) declaring what token is legal tender measured in that unit; and (3) enforcing debts and contracts payable in that token.

The point is that debts and contract prices must be expressed in terms of the unit of account while the token can be whatever the government chooses, and can be changed independent of the unit of account. In the U.S. the unit of account is the dollar, and the token is a dollar bill. When the U.S. established the dollar in 1792, the token was a gold coin of specified weight.

Any State with the power to tax can establish its own currency by declaring what token is to be legal tender. All modern States have adopted intrinsically worthless tokens for their currencies, known as fiat money. The State necessarily holds a monopoly on the production of fiat money. However it must issue enough to the public to enable it to pay its taxes, plus enough more for the economy to function efficiently.

The source of the State’s money tokens is normally its central bank. In the US, the tokens are carried as liabilities on the Fed’s balance sheet, backed by the financial assets bought from the private sector with those tokens. Those liabilities comprise the monetary base of the private sector. We will refer to those tokens as base money.

Base money acquires value because of its status as legal tender but more importantly because that is what the private sector must deliver in paying federal taxes. In effect base money is a tax credit. Those who have no tax liabilities will readily accept payment in base money because it is needed by so many others. The viability of base money ultimately depends on the government widely enforcing tax collection, and acting to maintain a modest rate of price inflation.

Banks as we know them today have two distinct roles. They are profit-seeking enterprises as well as depositories. Their profit-seeking activities include a variety of services. However we will focus on their role as intermediaries who provide a link between those with savings to invest and those in need of funds.

As depositories banks accept deposits, provide payment facilities, and issue cash on demand in exchange for deposits. They pay no interest on demand deposits and very modest interest on savings deposits. They also offer term deposits at higher interest rates because those deposits provide a more stable supply of funding to back their lending.

We will use the term banks to mean any financial institution that serves as a depository, such as commercial banks and thrifts. That does not include the subsidiaries of banks or bank holding companies, which cannot accept deposits but are permitted to engage in a variety of investment activities and to hold assets not allowed to banks themselves.

A fractional reserve banking system has two kinds of money, base money and bank money. The Fed creates base money when it purchases Treasury securities from the public. It pays by simply crediting the seller's bank with a deposit at the Fed, while the bank credits the seller with a deposit in his own account.

Base money is the definitive money of the nation, which means the government has no obligation to convert it into another form of asset. It comprises the cash held by the private sector and bank deposits at the Fed. All payments to and from the government require the transfer of base money. For example, when one writes a check to pay his taxes, his bank must surrender that much in reserves of base money to the Treasury for the check to clear.

Bank money refers to deposits in banks, all of which are claims on base money. The viability of bank money depends on the promise that it can be converted on demand into base money at par. Bank money is created when a bank issues a loan. It does so by simply crediting the borrower's account with a deposit. The bank must hold enough reserves of base money to meet the reserve ratio requirement on its demand deposits.

Bank money is the credit side of a balance sheet relation. Every dollar of credit in the form of bank money is matched by an equal amount of debt. For the borrower, a bank loan creates a credit (the deposit) and a matching debt (the obligation to repay the loan). For the bank, the loan creates an interest-earning asset (the loan contract) and an equal liability (the borrower’s deposit).

What is meant by the money supply in reference to the private sector? The term itself implies that a certain amount of money exists at any given time, even though the quantity may be unknown. In a fractional reserve system, there can be no meaningful measure of the money supply, as will be explained.

The Fed has its own arbitrary measures of the money supply which it once used to help guide its monetary policy decisions. It defines the money supply as the total cash in circulation and the deposit liabilities of banks and thrifts. At one time it set targets for the growth of the money supply. Now it largely ignores its own measures because it has found little correlation between them and its major policy objectives – limiting inflation and unemployment.

The Fed's definition of the money supply includes only what the non-bank sector holds. Thus the reserves of banks, i.e. vault cash and deposits at the Fed, are not included in the monetary aggregates, even though they are a part of the monetary base. That means when a bank makes payments to the public, it increases the money supply. When it receives payments from the public such as interest on loans, the money supply decreases.

An important shortcoming of the Fed's definition is that it ignores bank lines of credit which can be exercised at the discretion of the borrower. Firms often hold substantial lines of credit at their banks, which they can use on short notice. Likewise consumers hold lines of credit in their credit card accounts that are just as useful for purchases as checking accounts or the currency in their wallets. Lines of credit increase liquidity, which is ultimately what counts in terms of effective aggregate demand.

When someone uses a credit card in a purchase, he automatically expands the money supply as defined by the Fed. The seller receives a new deposit in his account, which increases the total of demand deposits in the banking system – until the buyer pays off the loan. Consumers who roll over their credit card loans rather than paying them off have increased the money supply on their own initiative by hundreds of billions of dollars. Thus the effective money supply is substantially larger and less measurable than the Fed's definition.

A private enterprise with sufficient financial capital may obtain a charter that permits it to accept deposits of base money from the public, and to issue loans in the form of bank money. When one deposits a check or cash in his account at a bank, he receives credit in exchange, namely bank money. We expect banks to redeem those credits for cash on demand and to honor checks written against those credits. Most of the money in use today by the private sector exists as credits issued by private banks.

When one pays by writing a check on his bank deposit, if the payee deposits the check in another bank, the payer's bank must transfer an equal amount of reserves to the payee's bank. Thus base money is the foundation of the bank money system.

In reality, base money itself is a form of credit. In the same way a contract can be viewed as a document or the agreement it represents, money can be viewed as a token or the credit it represents. Since credit for the holder is debt for the issuer, money can also be viewed as a token representing third party debt. In the case of base money, the third party is the Fed.

All base money originates with the Fed. For the most part, it is issued in exchange for securities the public bought from the Treasury with base money previously acquired from the Fed. This circular system of credit is difficult for some to understand, especially for those who think of money only in terms of the token itself rather than the credit represented by the token.

If base money is simply a form of credit backed by Treasury securities, which are another form of credit, then what assures the viability of base money, and what is the real basis of its value? The Fed's base money liabilities are backed by its assets in the form of Treasury securities which it previously bought from the public. But what prevents the real value of those Treasury securities from being diluted by deficit spending? As will be explained, the purchasing power of base money has very little to do with the amount of deficit spending. However it does depend in the long run on the cost to banks of acquiring base money, which the Fed itself controls.

Since base money is a monopoly of the State, the Fed must issue enough to avoid a shortage of what the public must use to pay its taxes. In practical terms, that means it must provide whatever reserves the banking system needs to ensure the liquidity of the payment system. When the Fed needs to increase aggregate reserves, it buys Treasury securities from the public and credits the sellers' banks with additional deposits at the Fed. Conversely the Fed sells Treasury securities to the public from its own portfolio when it needs to decrease aggregate bank reserves. Bank reserves are only a small part of the monetary base, but they play a key role because they are the grease that enables the bank credit system to function.

These transactions by the Fed are designed to balance supply and demand for bank reserves at the Fed's target interest rate on overnight loans between banks, otherwise known as the Fed funds rate. The Fed funds rate is the benchmark for all short-term interest rates. It has a significant influence on the amount of bank money issued, and thus the liquidity of the private sector. In controlling the Fed funds rate, the Fed necessarily relinquishes control of the amount of base money it issues. The private sector itself determines the net amount issued.

The Treasury spends out of its account at the Fed. It continually replenishes that account with transfers from its accounts in commercial banks where it deposits its receipts from taxes and the sale of bonds. These so-called Treasury Tax and Loan accounts in commercial banks are backed by deposits at the Fed, which are reserves of the banking system.

Treasury operations simply recycle base money previously issued by the Fed. The Treasury approximately balances its receipts from taxes and the sale of bonds against its spending in order to avoid large variations in the demand deposits of the private sector which could significantly affect liquidity. It targets a fixed balance in its account at the Fed in order to minimize variations in the aggregate reserves of the banking system. The Fed compensates for the variations by adding or draining reserves on a short-term basis through its open market operations.

If the private sector holds more base money than it needs, it will normally use the excess to purchase interest-earning Treasury securities, since base money earns no interest. The Treasury will always be able to recapture its deficit spending through the sale of securities, since it can pay whatever interest rate the market demands.

The interest rate the Treasury must pay to borrow is a market rate which is influenced by Fed policy. The short-term rate closely tracks the Fed funds rate due to arbitrage. Longer-term rates include a premium over the Fed funds rate which varies with inflationary expectations. Although many diverse factors affect those expectations, the Fed itself has considerable influence through its monetary policy decisions.

It is therefore up to the Fed to keep inflationary expectations within acceptable limits. By doing that well, it protects the purchasing power of base money, and ensures that interest rates on long term borrowing will not become so burdensome as to hinder economic growth.

The historical record shows no significant correlation between the amount of deficit spending and the inflation rate or interest rates. Most central banks now target a small positive inflation rate to provide a margin against a deflation trap. Deflation hurts aggregate demand by creating a money-hoarding psychology which is difficult to overcome, and may result in a prolonged recession. Under the gold-based system, the State's ability to counter inflationary and deflationary pressures was very limited.

London goldsmiths, originally operating as money changers, accepted coins and other gold objects for safekeeping for a fee, and issued receipts to the depositors. This has become known as warehouse banking. By the mid 17^th century, people found it more convenient to exchange the receipts rather than the coins when making payments among themselves. This facilitated trade within the economy, and the receipts themselves gradually became the accepted form of money. The goldsmiths found that people would rarely redeem the deposited gold for their receipts. Consequently they began issuing new receipts through lending, thereby creating receipts unbacked by gold deposits. Thus began the transition from warehouse banking to fractional reserve banking in England.

Unbacked receipts were the origin of the later banknote, a promissory note issued by a bank and payable in gold coin to the bearer on demand. Fractional reserve banking was an attractive means of expanding the money supply. However occasional over-lending by bankers created problems. When their promissory notes could not be fully honored, bank runs usually followed which sometimes resulted in serious consequences for the local economy.

A key difference in a modern fiat money system is the existence of the central bank. One of its roles is to act as lender of last resort. As the source of base money, it can lend whatever a bank needs to cover depositor withdrawals. If a bank is solvent but has a liquidity problem, it can borrow the funds it needs from the central bank.

The bulk of all money transactions today involve the transfer of bank deposits. A bank must hold reserves of base money in order to meet its depositors' cash withdrawals and to cover the checks written against their accounts. Reserves comprise a bank's vault cash and what it holds on deposit at the Fed, i.e. Fed funds.

When a depositor writes a check against his account, his bank must surrender that amount in reserves to the payee’s bank for the check to clear. Reserves are constantly moving from one bank to another as checks are written and cleared. At the end of the day, some banks will be short of reserves and others long. Banks redistribute reserves among themselves by trading in the Fed funds market. Those long on reserves will normally lend to those short. The interest rate on interbank loans varies with supply and demand.

The reserve requirement applies only to the bank's demand deposits, not its term or savings deposits. Thus when a bank depositor converts funds in a demand deposit into a term or savings deposit, he frees up the bank's reserves that were held against the demand deposit. The bank can then use those reserves in several ways. For example, it can hold them to back further lending, buy interest-earning Treasury securities, or lend them to other banks in the Fed funds market.

The Fed funds rate effectively sets the upper limit on the cost of reserves to banks, and thereby determines the interest rate that banks must charge the public for loans. The interest rate influences the demand for bank loans, and thus the net amount of bank money. Liquidity is an important factor in aggregate demand and inflationary pressure, which is why the Fed targets the Fed funds rate as its key monetary policy tool.

All depository institutions -- commercial banks and thrifts -- in the United States are subject to reserve requirements on customer deposits. The required reserve ratio depends on the amount of checkable deposits a bank holds. No reserves are required on the first $10.3 million. Between $10.3 million and $44.4 million, deposits are subject to a 3% reserve. Above $44.4 million they are subject to a 10% reserve. These breakpoints are adjusted annually in accordance with money supply growth. No reserves are required against time deposits or savings accounts.

Reserves are figured as the average held over a 14-day period, ending every second Wednesday. On any single day, a bank needs only enough to cover its customer's withdrawals. A bank may hold its reserves in any combination of vault cash and deposits at the Fed. As profit-seeking enterprises, banks try to keep their reserves close to the required minimum, since they earn no interest.

A bank loses reserves whenever it pays out cash or transfers funds by wire for its customers. Customer checks to pay out of town bills funnel back through the Fed and are charged against its reserves. A bank may also lose reserves when it advances loans or buys securities. Conversely a bank gains reserves when it receives new deposits.

A bank facing a reserve deficiency has several options. It can try to borrow reserves for one or more days from another bank; sell marketable assets, such as government securities; bid for funds in the money market, such as large CDs or Eurodollars; or as a last resort it can pledge collateral and borrow at the Fed’s discount window.

An active market in reserves acts to redistribute reserves to those banks that need them. However banks cannot create reserves themselves. If the aggregate demand exceeds the existing supply of reserves, the banking system as a whole has no alternative but to borrow reserves from the Fed.

There are many factors outside of the Fed’s control that influence the level of non-borrowed reserves. They include changes in currency holdings of the public, changes in the Treasury’s cash balances at the Fed, checking system float, and foreign central bank transactions. The Fed actively compensates for these variations by adding or draining system reserves as needed to avoid large fluctuations in their market price, i.e. the Fed funds rate. The growing demand for currency is the largest single factor requiring reserve injections.

The Treasury holds working balances at the Fed for making payments on behalf of the government. Drawing down those balances increases aggregate banking system reserves since it results in a transfer of funds to the banking system. In order to minimize variations in total banking system reserves due to its own spending, the Treasury targets a fixed balance of $5 billion at the Fed by transferring funds as required from its Treasury Tax & Loan accounts at commercial banks. TT&L accounts serve as collection points for receipts from taxes and the sale of securities, and are reserves of the banking system.

Many foreign central banks keep working balances at the Fed to execute their dollar-denominated transactions. Drawing down of those balances increases the reserves of depository institutions receiving payments. Transfers can sometimes result in significant increases or decreases in reserves, requiring offsetting open market operations by the Fed.

One of the main challenges to a bank is ensuring its own liquidity under all reasonable conditions. Liquidity for a bank means the ability to meet its financial obligations as they come due. Commercial banks differ widely in how they manage liquidity. A small bank derives its funds primarily from customer deposits, normally a fairly stable source in the aggregate. Its assets are mostly loans to small firms and households, and it usually has more deposits than it can find creditworthy borrowers for. Excess funds are typically invested in assets that will provide it with liquidity such as Fed funds loaned and U.S. government securities. The holding of assets that can readily be turned into cash when needed, is known as asset management banking.

Large banks generally lack sufficient deposits to fund their main business -- dealing with large companies, governments, other financial institutions, and wealthy individuals. Most borrow the funds they need from other major lenders in the form of short-term liabilities which must be continually rolled over. This is known as liability management, a much riskier method than asset management. A small bank will lose potential income if gets its asset management wrong. A large bank that gets its liability management wrong may fail.

The key to liability management is always being able to borrow. Therefore a bank's most vital asset is its creditworthiness. If there is any doubt about its credit, lenders can easily switch to another bank. The rate a bank must pay to borrow will go up rapidly with the slightest indication of trouble. If there is serious doubt, it will be unable to borrow at any rate, and will go under. In recent years, large banks have been making increasing use of asset management in order to enhance liquidity, holding a larger part of their assets as securities as well as securitizing their loans to recycle borrowed funds.

A bank run is an overwhelming demand for cash by a bank's depositors. With the advent of deposit insurance, bank runs by small depositors are largely a thing of the past. Insurance is limited to $100,000 per deposit, which provides complete coverage to about 99% of all depositors. But it covers only about three-fourths of the total amount of deposits because many accounts far exceed the insurance limits.

A large depositor assumes a risk and needs to know something about the bank's own balance sheet. However a healthy balance sheet does not eliminate all risk. Even if the depositor knows the bank has adequate liquidity, others may not. Large depositors must therefore be concerned about what others are likely to believe. A rumor about a bank, even though unfounded, can trigger a run that causes a solvent bank to fail.

The Fed acts as a depository for the Treasury as well as member banks. All government spending is paid out of the Treasury's account at the Fed. Whenever the government spends, the Fed debits the Treasury's account and credits the Fed account of the payee’s bank. The Treasury replenishes its Fed account with transfers from its commercial bank accounts where it deposits the receipts from taxes, and the sale of its securities.

On average, government spending does not affect the aggregate bank deposits of the private sector. The Treasury sells or redeems securities as required to balance its inflows against outflows. However short-term variations occur because receipts cannot be synchronized with spending.

The Treasury attempts to minimize disturbances to aggregate banking system reserves by maintaining a nearly constant balance in its Fed account. In effect, Treasury payments are transfers from its commercial bank accounts to the bank accounts of the public. Funds move in the reverse direction when the public pays taxes or buys securities from the Treasury. The Treasury must maintain an adequate balance in its commercial bank accounts to avoid having to borrow directly from the Fed. However it has no need for, and does not accumulate, balances in excess of its near-term payment obligations.

In a fully-backed system, every bank would have to hold reserves equal to the full value of its demand deposits. Reserves are held on deposit at the Fed and as vault cash. We will call the deposit at the Fed its reserve account. For its investment activities the bank would need additional funds at the Fed, which we will call its investment account. When the bank spends, the Fed would debit the bank's investment account and credit the reserve account of the payee's bank, while the payee receives a new credit in his own bank account.

A bank could no longer simply credit a deposit to fund a loan, as in a fractional reserve system. It would have to transfer funds from its investment account to its reserve account to cover the new deposit. Banks would have to hold enough cash to meet the withdrawal demands of their depositors. To acquire cash, banks would draw on their reserve accounts. That simply swaps one form of reserves for another without changing the total.

In a fully-backed system, the depository and payment functions of private banks would be basically clerical, and offer little opportunity for income other than sevice fees. Banks would have little interest in acquiring the reserves that come with new deposits. Their primary interest would be investing by leveraging their own capital. As we will see, there are good reasons for transferring the depository role to a single national depository run by the Fed.

In a single national depository, which we will call the National Bank, all deposits would exist as entries in a common computer system. Verifying balances and making payments could be done instantly, thereby eliminating checking system float with its logistic cost and complexities. The two-tier system of base money and bank money would be replaced by a one-tier system of base money alone, all created by the Fed. The entire money supply would consist of deposits at the National Bank and circulating cash. Since cash could be acquired only in exchange for deposits, the Fed would have total control of the money supply.

All deposits would be transaction deposits which pay no interest because they are the equivalent of cash. The National Bank would not offer interest-earning term deposits since they would be redundant with Treasury securities which are the functional equivalent of term deposits. Deposit insurance would be ended because they all deposits are risk-free. It is worth noting that when private banks no longer serve as depositories,the conept of bank reserves loses its meaning.

Henceforth we will assume a National Bank, operating through local branches which we will call banks. Its only role would be to accept deposits, execute payment orders, and provide cash in exchange for funds on deposit. The twelve Federal Reserve Banks should not be confused with the National Bank. However the Reserve Banks would provide the facilities and operate the National Bank.

Any economic entity, whether a financial firm, a non-financial firm, or an individual would be eligible to open a transqction account at a branch bank. Former banks would lose their depository role and become non-bank financial intermediaries, which we will call Financial Service Companies (FSCs). They could selltheir branches to the Fed and/or merge with existing intermediaries. FSCs already create most of the loans on which the economy depends. In a single depository system, they would create essentiall all of them.

The basic goal of monetary policy is to provide financial liquidity as needed to support a growing economy while limiting the inflation rate to some small value, typically about 2 percent per year. Monetary policy can be implemented in two ways: (1) control of the money supply growth rate, which leaves the money market interest rate as a residual, or (2) control of the money market interest rate, which leaves the money supply as a residual. The money market rate is the interest rate at which major FSCs lend money to each other on a short-term basis, say 30 days. It is analogous to the London Inter-Bank Offer Rate (LIBOR) on 30-day Eurodollar placements.

Some economists have advocated option (1) with a fixed growth rate of the base money supply consistent with the potential growth rate of the real economy plus a small inflation rate. A key objection to this is its inflexibility. There are certain to be significant external shocks to the economy from time to time. The negative impact of such shocks could be lessened with adjustments in interest rates, which a fixed money growth rate policy would not allow. Furthermore the potential growth rate exhibits long term variations for a number of reasons, which makes determination of an optimum fixed rate problematic.

Option (2) is adopted here because it provides the flexibility to deal with unexpected conditions. It is also more effective in limiting interest rate volatility. Firms cannot plan efficiently if interest rates vary sharply and unpredictably. Selecting the optimal interest rate at any given time is not easy, and mistakes have been made. However a great deal of experience now exists in many different economies which support the overall effectiveness of the interest rate targeting option.

The Fed steers the money market interest rate toward the target rate through its open market operations. When the interest rate rises above the target rate, the Fed purchases securities in the open market to increase the aggregate supply of deposits. When the interest rate falls below the target rate, the Fed sells securities. In this way it continually adjusts the supply of loanable funds to meet the demand at approximately the target rate.

In the fractional reserve system, the Fed controls the money market interest rate by operating directly on the reserves of banks. Since the loanable funds supply is a much larger pool of money, interest rate control in a fully-backed system will be less responsive to the Fed's open market operations, and will therefore require far more funds.

The normal growth bias of the economy creates the need for an expanding money supply. In order to maintain control of the overnight lending rate, on average the Fed must purchase securities from the public to support the demand. That means the Fed’s portfolio of securities will continue to grow with time.

FSCs perform a vital role in the economy. They pool the savings of investors and make them available to borrowers. They sell financial instruments as diverse as debt securities, mutual funds, and insurance policies. In issuing loans they increase the amount of private sector debt. In principle there is no limit to how large private debt can grow. In practice private debt should not exceed what borrowers can safely cover with their existing wealth or with their own income as payments become due.

Even though the amount of debt increases as a function of demand, the money supply remains unchanged until the Fed acts to change it. However the Fed has to provide whatever new money is needed to hold the interest rate in line with its target rate, as long as its monetary policy is based on the interest-rate priority option. In the current fractional reserve system, most of the private sector debt is generated by FSCs rather than banks. In a fully-backed system, virtually all of the private sector debt would be generated by FSCs.

An FSC in good standing may unconditionally borrow from the Fed, just as banks do now. The interest rate is set 100 basis points above money market target rate. This large spread means that FSCs will use the lending facility mainly to cover short term cash flow problems, and not as a source of funds to invest. The loan must be collateralized by the borrower with Treasury securities. It may be rolled over indefinitely as long as the borrower has sufficient funds on deposit to pay the interest and to cover any change in the market value of the collateral.

FSCs eligible to borrow from the Fed should be limited to those in the “too large to fail” class. Other FSCs and non-financial firms with good credit should be able to borrow in the money market, just as households with good credit should be able to borrow from a FSC offering retail loans.

The Treasury sells securities to the Fed only to roll over the maturing securities in the Fed's portfolio. Otherwise all new Treasury securities are sold to the public. This ensures that the yields reflect the open market rate. The principal source of income for the Fed is the interest it earns on its Treasury securities. Periodically, the Fed remits any income in excess of its operating costs to the Treasury.

In a fractional reserve system, the Treasury holds most of its funds in commercial banks in order to facilitate the Fed’s control of the Fed funds rate. In a fully-backed system with only Fed branch banks as depositories, the Treasury would hold all of its funds at the Fed. It must manage its cash flow to maintain its Fed account at a fixed level on average in order to minimize variations in the private sector money supply. That means the Treasury must cover its deficit spending in a timely manner with the sale of its securities.

Cash flow is a constant concern for financial institutions. FSCs borrow to lend at a profit, but they differ widely in the degree to which they mismatch the maturity of their assets and liabilities. Mismatching creates a potential cash flow problem. Most FSCs borrow short and lend long, expecting to roll over their short term liabilities on a continuing basis. If their credit became suspect for any reason, they might be unable to refinance at all. There is nothing about a fully-backed system that would eliminate excessive risk taking by FSCs.

The risk of systemic failure arises from the cascading of liabilities of one FSC on another. A default by a major borrower could cause defaults by earlier lenders. To minimize this danger, capital adequacy requirements should be imposed on all FSCs considered "too large to fail". The required ratio of capital to risk-adjusted assets should be an increasing function of the mismatch in their maturities. This is similar to the capital adequacy requirements now imposed on banks, based on the Basel accord. The details of the formula need to be worked out, and will require careful study to ensure its effectiveness and practicability.

In a fully-backed system there is no need for deposit insurance. All deposits represent base money on deposit at the Fed. Government insurance should not be provided on the financial instruments offered by the FSCs. Bond rating agencies can expand their domain to rate the creditworthiness of FSCs as institutions. That would be of great value to investors, but more importantly it means that FSCs would tend to avoid practices that reflected badly on their creditworthiness in the competitive business of lending for a profit.

There should be less temptation for FSCs to advance loans for purely speculative games. However the risk of too rapid growth through borrowing and lending would remain so capital adequacy rules should be extended to major FSCs as they are now to banks.

It eliminates the need for deposit insurance and the moral hazard that can foster.

With the Fed operating as the sole depository, payments can be cleared without delay, eliminating the nuisance of checking system float, and significantly reduce associated costs.

It should end sterile games that banks play to get around the fractional reserve requirement, such as overnight sweep accounts..

Since money earns no interest, firms and households will hold the minimum required to meet their liquidity needs. Holding money in excess of near-term needs represents a loss in purchasing power equal to the inflation rate, which has averaged about 3 percent over the long term. Alternatively, it represents an opportunity cost which varies with the money market interest rate.

If the fully-backed system were working in the U.S. today ( Aug 2008), a rough estimate of the base money supply would include the currency in circulation ($900 billion), total demand deposits and other checkable deposits ($700 billion), overnight sweep accounts ($800 billion), and the eurodollar liabilities of US banks ($600 billion). The latter figure is a guesstimate because there are no reliable figures for the total.

The estimated total base money supply would therefore be about $3,000 billion. That is an increase of about $2,000 billion from the current amount, which means the Fed would have to monetize that much of the $5,800 billion in Treasury securities held by the public.

The transition would not be easy. It would have to be carefully planned and phased in slowly enough to give all parties adequate time to make the necessary adjustments. A logical way to proceed would be to gradually increase the reserve ratio requirement on existing depositories until it reached 100 percent. By that time, the interest-earning deposit liabilities of banks should have been paid off, leaving only demand deposits fully backed by reserves. To acquire the funds to pay off those deposits, banks would likely have to mature or call the bulk of their loans.

In setting up a single national depository, i.e. a National Bank, all demand deposits of banks and the reserves backing them would be transferred to the National Bank after 100% reserves had been achieved. Any excess reserves would be credited to the private banks' accounts at the National Bank, which together with their remaining investments would comprise their net financial worth. Private banks would then evolve into FSCs, or they could sell out to existing FSCs. Without the depository role, they would no longer need the same number of branch offices. The Fed would probably offer to buy some of them in setting up its own depository branch offices.