II Systemic Financial Risk in Payment Systems
Author:
Mr. D. F. I. Folkerts-Landau https://isni.org/isni/0000000404811396 International Monetary Fund

Search for other papers by Mr. D. F. I. Folkerts-Landau in
Current site
Google Scholar
Close

Abstract

At the core of current international monetary arrangements lies an interlocking network of national and international payment systems that facilitate the exchange of funds associated with almost all international trade and financial transactions. These payment systems have been an almost invisible component of the international monetary system because, to a large degree, they have functioned smoothly and efficiently. However, during the past two decades, the growing size and integration of major financial markets has sharply increased the volume of transactions both within and across national payment systems. As a result, private and official sector observers have expressed concerns about whether existing institutional arrangements in the major payment systems can both efficiently cope with the new volume of transactions and effectively manage the risks created by such factors as counterparty failure and liquidity crises. An extended disruption in the largest payment systems would clearly have a highly adverse effect on international trade and financial flows.

At the core of current international monetary arrangements lies an interlocking network of national and international payment systems that facilitate the exchange of funds associated with almost all international trade and financial transactions. These payment systems have been an almost invisible component of the international monetary system because, to a large degree, they have functioned smoothly and efficiently. However, during the past two decades, the growing size and integration of major financial markets has sharply increased the volume of transactions both within and across national payment systems. As a result, private and official sector observers have expressed concerns about whether existing institutional arrangements in the major payment systems can both efficiently cope with the new volume of transactions and effectively manage the risks created by such factors as counterparty failure and liquidity crises. An extended disruption in the largest payment systems would clearly have a highly adverse effect on international trade and financial flows.

This study examines both the nature of systemic risks in payment systems and the policies being implemented by the authorities in major countries to manage or curb them. The first section of this study briefly reviews the role, evolution, and public policy issues associated with modern payment systems. The next two sections identify the main characteristics of systems for clearing and settling payments and examine the financial risks in payment systems. Then policy initiatives aimed at containing payment system risks are discussed, and finally the issues are summarized.

Payment Systems: Role, Evolution, and Public Policy Issues

Role of Payment Systems

Most economic transactions in market economies involve an exchange of goods, services, or securities for money. In modern economies, money largely consists of bank liabilities; hence, the role of a payment system is to effect the transfer of bank liabilities among transactors.1 Although the payment systems in the major industrial countries share a common objective, their institutional arrangements differ as a result of differing patterns on the use of checks versus electronic fund transfers, the use of post office accounts versus bank accounts, and the relative importance of large wholesale payments.2 Retail payments tend to be large in volume of transactions but small in size, while wholesale payments are by their very nature large in size and are executed electronically in all major payment systems.3 Moreover, retail payments made by checks tend to be free of systemic risk since the liability for payment falls on the payor only and the failure of a single payor or a group of payors to make payments is unlikely to affect the depository intermediary.

In contrast, most wholesale payments involve exchanges between large institutions that arise from transactions by their customers in securities markets. Banks are typically the main participants in wholesale payments systems.4 Banks play an important role both because of their ability to establish efficient payments arrangements and because of their direct access to “good funds” (reserves at the central bank) for payments. Such good funds constitute the core of any payment system because they are available for the use of the owner to make payments under all market conditions (especially during market crises).

At first glance, the payments side of a transaction appears to be straightforward—the payor instructs his bank to transfer an amount of funds agreed upon to the bank of the payee. But closer inspection reveals a number of difficulties that could prevent a successful transfer from taking place. First, there is the problem of how best to achieve a simultaneous exchange of goods or securities for money so as to minimize the risk that one party to the transaction could renege on his payment or his delivery obligation, while the other party is fulfilling his side of the transaction. This risk can be minimized by closely tying the delivery of securities or goods to the receipt of good funds by the seller. The payments system is therefore closely linked to the system of transferring securities and legal arrangements for securing title to goods.

Second, there is the problem of how to transfer funds from one bank to another. In most systems, this problem is ultimately solved by having the central bank transfer central bank liabilities (that is, high-powered money) from the reserve account of the sending bank to the reserve account of the receiving bank. However, this transfer of funds is often achieved indirectly by using correspondent bank relationships or clearinghouse arrangements. A clearing or money center bank may clear payments among a number of smaller correspondent banks that have accounts with the clearing bank. Under this arrangement, each correspondent bank would only have to transfer (or receive) the net amount of his payment to and receipt from the other correspondent banks. In addition, groups of banks in some countries have created clearinghouses that net payments among the banks before achieving final settlements on the books of the central bank, for example, CHIPS in the United States and giro clearinghouses in Germany.

Third, the issue arises of how to deal with a payments failure. It is common practice in most countries for banks to send and receive payments instructions from a clearinghouse or central bank throughout the day and to achieve final settlement only at the end of the day or at the beginning of the next day by transferring the net amounts owed/due on the books of the clearinghouse or the central bank. If an insolvent bank fails to have sufficient good funds at settlement time, the issue of which parties bear the losses is typically resolved by either (1) unwinding all the payment instructions sent out by this bank during the day (thereby imposing the losses on the counterparties to the bank’s transactions); (2) letting the settlement stand and imposing a cooperative assessment of losses on the members of the clearinghouse or payments system; or (3) providing temporary public sector credits and resolving the sharing of loss through legal channels.

Evolution of Payment Systems

Structural changes in major payment systems have reflected the evolving transaction requirements of modern banking systems. To an important degree, the growth of modern fractional reserve banking systems was initially stimulated by reductions in the transaction and payment costs that could be obtained by substituting promises to pay of banks for commodity money or currency.5 To obtain these cost savings, however, the depositor had to relinquish good funds (such as currency) and become a general creditor of the bank. Further gains in payment efficiencies were made by banks through the holding of interbank credit and debit balances on each other overnight or for longer periods, which allowed banks more time to adjust to payments shortages.

However, the bilateral clearing and settling of checks between large banks meant that messengers went with bundles of checks from bank to bank on which they were drawn. Such bilateral methods resulted in long periods between settlement during which large amounts of “float” (free credit to the bank on which the check was drawn) would accumulate. The next stage in the evolution of banking and payments systems, therefore, was the displacement of bilateral bank payment relationships by multilateral clearinghouses. In the United States, for example, such clearinghouses first emerged during the middle of the nineteenth century in the form of private cooperative arrangements to economize on check collection.6 The clearinghouse member banks brought checks to the clearinghouse at a fixed settlement time, and after the checks were netted, each member’s net credit or debit position against the clearinghouse was computed and banks with debit positions were required to deliver government currency or coins, which were then passed on to the net creditor banks. The later practice of keeping reserves at clearinghouses facilitated an overnight interbank market that produced a more efficient distribution of reserves among banks and allowed banks to reduce their reserve holdings. The period of float was also reduced to a period of hours.7

The clearinghouse traditionally provided another benefit, namely, some payment finality. Clearinghouse members usually agreed to cover the net debit positions of a failed member bank out of an assessment on their transactions. The clearinghouse then became a general creditor of the failed bank. Thus payment finality meant that the depositor of the check was at least partially insured against the failure of the paying bank from the time of deposit to the time when his account was credited. Such clearinghouse corporations thus allowed member banks to delegate the monitoring of each bank’s solvency to the clearinghouse. In return, the clearinghouse restricted membership, imposed capital requirements on members, and held periodic inspections. Nonmember banks cleared their checks through correspondent banks.

The clearance of payments required clearing banks to offer lines of credit to their correspondent banks, which meant that clearing banks had to acquire expertise in monitoring and managing interbank credit, which frequently arose in the clearing process on short notice and without the safety of collateral. Since clearing banks usually were best able to evaluate the creditworthiness of correspondent banks, they became the main supplier of short-term liquidity for the payments system.

Public Policy Issues

The growing involvement of central banks was the final step in the development of modern banking and payment systems. For payment systems in a fractional reserve system to be stable and to work efficiently, checkable deposits must remain convertible into currency. If a clearinghouse restricted conversion of deposits into currency, deposits would sell at a discount in terms of currency. Thus the prospect of restricted convertibility could cause forward-looking depositors to convert deposits immediately into currency, thereby precipitating a liquidity crisis.

The ability to create currency through the open market purchase of securities or direct lending against eligible collateral has allowed central banks to offset these liquidity shortages and thereby helped maintain the exchange rate between bank deposits and currency.8 In addition, central banks have also served as clearinghouses where banks can hold their clearing balances. In providing short-term liquidity to banks, however, central banks have taken on a certain amount of credit risk as the cost of providing a more efficient payments system. In effect, the public as a whole has assumed some of the credit risk inherent in bank assets that serve as collateral for central bank lending in return for having an efficient payments system.

The role of the central bank in maintaining the stability of payment systems has raised important policy issues. The real economic costs of a breakdown in the wholesale payments system provide a key rationale for central bank intervention. Nevertheless, such intervention implies a policy of supporting clearing or money center banks that are the main participants in the payments system—creating the perception in some countries that some banks are “too large to fail.” As a result, depositors and other creditors of these institutions may tend to underplay the riskiness of the institution’s assets. Moreover, central banks have during crises sometimes not only provided general liquidity but have also indicated support for certain large institutions. The problem then is that such an extension of the official safety net can encourage the very kinds of excessive risk-taking that ultimately transpose the safety net into a large contingent liability of the official sector.

This problem has been magnified because large clearing or money center banks have supplied liquidity to facilitate not only payments transactions but also to finance short-term positions.9 Liquidity is supplied to the nonbank financial sector through revolving lines of credit that are designated to “back” the issue of securities, such as commercial paper. These revolving lines of credit assure the lender that he has access to good funds if the borrower is unable to roll over the security at maturity. Moreover, banks finance the inventory of securities dealers and other market makers through repurchase agreements. As a result central bank liquidity support for large banks implicitly extends official support to other institutions that rely on the liquidity provided by banks.

In recent years, there has also been a growing concern that developments in the major financial markets—increased international integration, higher volatility of asset prices, growth of derivative markets, and, above all, substantially larger trading volumes in all markets—are severely testing the adequacy of the existing infrastructure for clearing and settling large-value payments among major international financial institutions. The settling of payments, by the delivery of good funds at periodic, usually daily, intervals is a key test of the solvency of financial institutions in the international financial system. An international financial crisis, if it occurred, would most likely first manifest itself through the inability of a financial institution, or a group of institutions, to settle its obligations in one of the major payment systems. The fear is that such an event would cause an inadequately prepared payments system to “freeze”—become unable to effect payments among institutions. Such an inability to settle payments could then be expected to lead to a severe liquidity shortage, as healthy institutions—not having received payments expected at settlement time—might be unable to settle their own payments obligations.

Without central bank intervention, such a liquidity crisis could easily lead to a loss of confidence in depository institutions, which in turn could precipitate multiple failures of otherwise healthy financial institutions.10 As a result, major central banks have reassured financial markets of their liquidity support during times of stress. However, the sheer size of average daily payments flows—$1.4 trillion in 1988—through the domestic and international U.S. dollar wholesale payments system and the difficulties experienced in settling trades and payments following a computer breakdown at a single clearing bank in New York in 198511 and during the October 1987 equity price downturn have contributed to a sense of unease. In fact, some observers believe “that the greatest threat to the stability of the financial system as a whole during the October stock market crash was the danger of a major default in one of the clearing and settlement systems.”12

In addition, the growth of private, frequently offshore, payment netting arrangements has raised questions not only about systemic risk, but also about monetary control. If a central bank accommodates an abrupt increase in the demand for central bank money during a financial crisis, it could make it more difficult to control the expansion of the monetary base over the medium term. But failure to increase the supply of central bank money during times of liquidity shortages may result in a systemic liquidity crisis. Central banks have sought to improve the trade-off between control of the monetary base and the stability of the payments system by regulating and supervising private payments clearing arrangements. However, when private settlement systems operate outside the jurisdiction of the central bank, such supervision is more difficult to implement. Nonetheless, problems in an offshore clearinghouse might adversely affect the stability of a national payments system. For example, if private financial institutions set up a clearinghouse for dollar-denominated payments outside the United States, final settlement of payments would still ultimately involve clearing payments through the U.S. wholesale payments system. In this situation, problems in this offshore clearinghouse might adversely affect the stability of the U.S. payments system.

Such growing interdependence between national and offshore payment systems has led financial authorities in the major countries to undertake an extensive program to strengthen wholesale payment systems. First, steps have been taken to enhance the ability of payment systems to withstand operational and liquidity shocks and to allow for an orderly completion of the settlement process in the event of insolvency of a single institution or group of institutions. Second, a strengthening of the capital positions of international money center banks, as agreed in the recently implemented Basle Agreement, was aimed at improving confidence in the ability of these key players to withstand adverse credit or liquidity shocks and thereby enabling them better to fulfill their settlement obligations. Before considering these policy measures in detail, it is useful first to examine some examples of major payments systems to identify more specifically what characteristics generate liquidity and credit risks.

Main Features and Examples of Payment Systems

Main Features

Most wholesale payment systems consist of (1) a central bank that settles payments among a group of clearing banks via their reserve accounts (such as the Fedwire system in the United States); and (2) various private clearinghouse arrangements among subgroups of banks (such as the CHIPS for international dollar payments or regional or giro clearing systems in Germany). The main characteristics of these payment systems that are relevant for a discussion of systemic risk and public sector credit risk are gross or continuous settlement systems and net periodic settlement systems (Tables 49 of Appendix I summarize the key characteristics of major national payment systems).

In gross or continuous settlement systems, each payment instruction to the clearinghouse (central bank or private clearing corporation) results in an immediate debit for the sender and credit for the receiver in the settlement accounts that the sending and receiving banks hold with the clearinghouse. If gross or continuous settlement systems do not allow overdrafts against the clearinghouse, as is the case in Switzerland, payment instructions can be executed only if the payor has a sufficiently large credit in his account. If the clearinghouse allows overdrafts, as in the Fedwire system in the United States, it is necessary to determine the length of the period at the end of which the overdraft has to be eliminated. In this case, the clearinghouse is exposed to the intrasettlement period credit risk arising from these overdrafts.

In net periodic settlement systems, payment instructions are accumulated by the clearinghouse over a period of time, and only net debits and credits are entered into the settlement accounts of the members at the end of the settlement period. An important determinant of systemic risk in these systems is the presence or absence of settlement finality. Under a regime of settlement finality, any payment instruction received by the clearinghouse is irrevocably executed even if the bank sending the message defaults.13 Settlement finality precludes the clearinghouse from unwinding payment instructions if one or several members are unable to supply good funds at the end of the clearing day to settle their debit balances. Moreover, the nondefaulting participants of the settlement system are obliged to cover the shortfall at settlement. Liquidity crises are therefore avoided, while the claim against the defaulting institutions is resolved through legal recourse.

The risk-sharing rules adopted by the clearinghouse have to be explicitly formulated and must define the position of the defaulting institution vis-à-vis the remaining participants. The longer the settlement period, the greater the credit risk to which the clearinghouse is exposed. Most payment settlement periods are therefore one day or shorter. On the other hand, the shorter the settlement period, the greater is the need for institutions to hold costly reserves to be able to settle payment balances.

Gross settlement systems automatically achieve settlement finality, since each individual payment instruction is executed without netting, and thus unsettled balances do not accumulate. Furthermore, most gross settlement systems also have payments finality, that is, the payment instruction is irrevocably executed and cannot be revoked.

Examples of Major Payment Systems

The most prominent example of a gross payments system with payment finality and intrasettlement period overdraft facilities is Fedwire, the world’s largest wholesale payment settlement mechanism. Fedwire is the U.S. Federal Reserve’s nationwide wire system for transferring funds and U.S. Government securities among foreign and domestic depository institutions operating in the United States. The depository institutions participating in Fedwire operate through the Federal Reserve Banks in their districts. As of January 1989, Fedwire had 11,398 active participants, of which 6,163 participated on-line. During the first quarter of 1989, the average daily transaction volume was 231,000, while the average daily value was almost $700 billion (excluding CHIPS net settlement). Direct access to Fedwire is restricted to depository institutions, while other financial and non-financial firms, securities firms, and insurance companies can gain access to the system only on terms and conditions set by their depository institutions.

Fedwire payments are made by debiting and crediting reserve accounts maintained by the respective depository institutions at their Federal Reserve Banks. The Fedwire payment is finally and irrevocably paid when a reserve bank sends the payment message to the receiving bank. Funds are immediately made available to the receiving customer (Federal Reserve Regulation J). Thus payment and settlement are final. The Federal Reserve Bank will execute a payment instruction even if it leads to a debit balance, and the execution of the payment does not depend on the account of the sender. If the sending bank failed while in overdraft, the risk would be borne by the Federal Reserve Bank, which would then become a general creditor of the failing bank. Such an overdraft must, however, be settled by the end of the day; hence, the term “daylight overdraft” indicates recourse to the federal funds market if necessary.

Fedwire intraday overdrafts occur when a depository institution’s outgoing Fedwire payments exceed the sum of its opening balance and its incoming Fedwire credits. The precise measurement of daylight overdrafts requires rules to determine when, during the day, debits and credits to a depository institution’s account at a reserve bank occurred. For Fedwire transactions the timing is clear, since they are considered to be final payments when the receiver of funds is advised of the credit. Since intraday reserve balances do not count toward meeting reserve requirements, banks are then primarily concerned about their reserve balances as of the end of the day—as daylight overdrafts do not carry a charge. A bank generally brings its reserve position up to that required by the end of the day through borrowing in the federal funds market or reduces its reserves to required levels by lending in this market. These funds are then returned the following morning.

U.S. Government securities are also transferred among banks over Fedwire. Each Federal Reserve Bank maintains ownership records of the securities in its computer system. Depository institutions can transfer securities held in their name to other institutions through a system of book entries. Such a transfer can be arranged either in conjunction with a transfer of reserves of equal value or as a separate transaction. Such security transactions also contribute to daylight overdrafts. Since reserve accounts are typically debited when book entry securities accounts are credited, a few clearing banks that specialize in transactions with dealers in government securities generate a large share of total daylight overdrafts of bank reserve accounts. For example, in the second quarter of 1988, four clearing banks accounted for about 70 percent of the daylight overdrafts attributable to transactions in book entry securities. Dealers in government securities maintain book entry securities accounts and demand deposit accounts with clearing banks but have no direct access to the Fedwire book entry system. They generally hold large inventories of securities during the day to meet the anticipated demands of their customers, but sell most of their securities by the end of the day through repurchase agreements (“repos”) to minimize the cost of holding the inventories. Investors who enter into these repurchase agreements own the securities overnight and resell them to dealers early the next day.14

Fedwire daylight overdrafts averaged between $60-65 billion a day in 1989, with book-entry-related overdrafts accounting for about 60 percent of all Fedwire peak intraday overdrafts. The six largest clearing banks account for about two thirds of all book-entry-related daylight overdrafts, while the ten largest clearing banks account for approximately 80 percent of all such overdrafts. Transfers of book entry securities over Fedwire averaged $312 billion a day in 1987.

The largest net settlements system without receiver finality is CHIPS (the Clearing House Interbank Payments System), the international dollar payments system. CHIPS is a private payments network owned by the New York Clearing House Association, with about 140 participating domestic and foreign banks, of which 22 are settling banks (banks that settle daily transactions on a net basis for their own account and as correspondents for nonsettling participants). Settlements among the 22 settling banks are made at the end of the day through the Fedwire system. CHIPS was the first private clearinghouse arrangement that permitted a real-time exchange of electronic payment information with net balances being at first settled the next morning. In October 1981, CHIPS began same-day settlement through a special account at the New York Reserve Bank. Hence, overnight and weekend float disappeared from the CHIPS system, leaving only daylight float. Most payments on CHIPS are associated with foreign exchange and Eurodollar transfers. The payments volume is about $700 billion a day, about the same as on Fedwire during the fourth quarter of 1988. Peak volume was reached on November 14, 1988, at $1.2 trillion. The peak intraday net debit position on CHIPS of about $45 billion a day has been smaller than the Fedwire daylight overdrafts of about $60-65 billion a day.

In 1986, the Federal Reserve Bank of New York undertook a special survey that focused on CHIPS and Fedwire to ascertain the nature, timing, and composition of payments on a single day (June 4, 1986) by sampling individual transactions (see Table 1). On that day, there were 120,000 CHIPS transactions with an aggregate value of $432 billion, and there were 56,000 Fedwire payments with a dollar value of $265 billion (these were payments originating in the New York Federal Reserve district only).15 The sample showed that Fedwire accounted for virtually all payments related to transactions for securities purchases/redemption/financing and for purchases and sales of federal funds, while CHIPS handled payments for almost all foreign exchange transactions. Overlap between the two systems occurred in the categories of payment-related bank loans, commercial and miscellaneous transactions, settlement, and Eurodollar placements. Foreign exchange and Eurodollar placements made up more than 80 percent of CHIPS dollar volume, while accounting only for 10 percent of Fedwire transactions.

Table 1.

Estimated Aggregate Transactions by Survey Category by Wire System1

article image
Source: Federal Reserve Bank of New York, Quarterly Review (Winter 1987–88).

The sample consisted of 13 banks accounting for 48 percent of all CHIPS payments on June 4, 1986 and of 9 banks accounting for 76 percent of all Fedwire payments on that day.

In millions of dollars.

Of the 265 foreign-based depository institutions that have a banking presence in the United States, 91 are CHIPS participants (two thirds of all CHIPS participants). Virtually all major depository institutions based in the United States are Fedwire participants, but fewer than 50 of these are represented on CHIPS. On the sample day, $270 billion of payments of Fedwire were related to securities transactions, while only $1 billion of CHIPS transactions were related to securities business. Bank loan transactions were generally low in frequency and value, indicating that much of this business is done on the bank’s own books. Almost all federal fund transactions occur over Fedwire. CHIPS usage is largely confined to foreign bank customers without direct access to Fedwire. The adjustment of correspondent balances constituted the dominant purpose for settlement transactions on both CHIPS (67 percent of transactions, 84 percent of dollar volume) and Fedwire (81 percent of transactions, 61 percent of dollar volume). CHIPS settlement transactions over Fedwire represent 5 percent of the settlement transactions on Fedwire, but more than 30 percent of the dollar volume. CHIPS transacted $23 billion related to foreign transactions, while Fedwire transacted $250 billion of foreign exchange transactions.

A net settlement system that nets payments on a continuous basis avoids, or at least reduces, the systemic risk associated with the possibility of a participant failing to settle at settlement time. An example of such a system is the FXNET, a U.K. limited partnership owned by subsidiaries of 12 banks, which was developed largely to deal with foreign exchange transactions. For any given value date and currency, the successive trades between two FXNET participants are continuously netted throughout the day. Such netting is bilateral, that is, it occurs between any two participants in FXNET. Thus, the decisions regarding creditworthiness exposure and risk management are under the control of each counterparty. Such netting is done by novation, that is, each trade is folded into the previous obligations and creates a new net position. This new obligation represents the only remaining binding bilateral currency payment obligation for all of the previous trades. Each counterparty either makes or receives one payment for each currency dealt on each settlement date. The exchange of two payments is no longer required for each pair of currencies dealt.

Netting by novation transforms a FXNET obligation into a stream of net payments over all forward dates dealt. Such netting reduces settlement risk, since the average amounts to settle are reduced and the volatility around the average is reduced. Without an effective netting agreement, a liquidator could choose not to honor the solvent party’s profitable contracts but could honor all the contracts that were profitable for the liquidating firm. Under the netting scheme, the liquidator has a claim only to net credits due to a liquidating firm and can default only on net debits. This feature greatly reduces systemic risk since it reduces the exposure of the overall system to the defaulting participant.

An example of a gross settlement system with payment finality that does not extend intraday overdraft to the senders of payments is the Swiss Interbank Clearing System (SIC).16 The SIC is a gross settlement system with payments being settled individually on reserve accounts. It uses queuing to prevent daylight overdrafts on these accounts, operates 24 hours on bank working days, and has a capacity of approximately 600,000 payments a day. Payments are irrevocable and final. Payment orders that would lead to an overdraft are automatically queued until sufficient funds have accumulated from incoming payments and are then automatically released for settlement on a first-in/first-out basis. This centralized facility relieves participants of having to synchronize incoming and outgoing payments to prevent overdrafts on reserve accounts. The queued orders may be canceled by the sending bank at any time. This rule was designed to minimize incentives for participants to make use of payments held in the queue.

A value day starts at around 6:00 p.m. and ends at around 4:15 p.m. of the following bank working day. The 24-hour operations feature allows the coordination of settlement of foreign exchange transactions, in which currencies of countries located in different time zones are involved. Further globalization of financial markets might lead to increased use of this feature in other settlement and payment systems. The SIC participant has real-time access to all data entered into this system relating to his account, that is, he can monitor settled incoming and outgoing payments as well as payment orders stored in the waiting queue and pre-value data file, as well as the actual balance of the reserve account. The value of payments reached Sw F 140 billion on an average day in April 1989—about 50 percent of annual GNP. The value of payments drops to approximately 10 percent of the average on U.S. bank holidays, indicating that foreign exchange transactions are the major source of large payments. Payments of about Sw F 1 million or more comprise 97.9 percent of total value.

Since a change of liquidity regulations in January 1988 in Switzerland made reserve account balances voluntary, all reserves became excess reserves. As a result, the balances held by SIC participants decreased from Sw F 7.5 billion to Sw F 2.6 billion in April 1988, or to about one third of their original level. The daily turnover ratio—the relation between daily payment value and reserves—has increased from 12 to 54 on an average day. Almost one third of payments are made and settled within ten minutes of being validated, while nearly two thirds are made within two hours, and only 2.5 percent of payments are in the queue for more than five hours. While the Swiss SIC system has eliminated daylight overdrafts, it is nevertheless subject to the risk of having a payments grid lock, a situation in which no payments move over the system because the accumulated credit balances are too small.

There are indications that, if the composition of the payment stream is not changed by subdividing very large payments (those in excess of Sw F 0.5 billion), further reductions in the reserve account balances may increase the frequency of payment grid locks. Grid locks are currently resolved at the end of the day with funds raised in the market or through collateralized lines of credit with the Swiss National Bank at a penalty rate. Alternatively, payments pending in the queue may remain unexecuted. Moreover, it is unclear whether sufficient incentives exist to prevent participants from reducing their holdings of reserve account balances to a level where major grid locks become likely.

Risk in Payment Systems

There are four basic parties to each transaction in a payments system: the payor originating the transfer, the sending bank that transmits the payments message of the payor, the receiving bank that acts on behalf of the payee, and the payee. Credit risk arises from the possibility that one of the parties in the chain of transactions defaults on its obligations. For example, the sender could initiate a transfer with his sending bank without having sufficient funds in his account to cover the transfer. The sending bank incurs credit risk if it transmits the payments message before the sender supplies the covering funds. Second, the sending bank may fail to provide funds to the receiving bank at settlement. Finally, the payee runs the risk that the receiving bank will not make the funds available. These three risks—sender risk, receiver risk, and settlement risk—are to be found in most payment systems. If the clearinghouse operates under settlement finality, the credit risk of the sending bank is distributed over the receiving banks according to the loss-sharing formula adopted by the clearinghouse.17

There is also systemic risk, which occurs as an outgrowth of settlement risk (see Appendix II). The failure of one participant to settle deprives other institutions of expected funds and in turn prevents these institutions from settling. Thus, although a participant does no business directly with a failed institution, chains of obligation may make it suffer because of the impact that the failed institution has on an intermediate participant’s ability to settle, that is, the cost of settlement failure reaches beyond the exposure of the credited bank to the failing bank. While it is generally not difficult to identify credit risk in a payments system, it is difficult to identify systemic risk properly.

While a private or public payments system with settlement finality, such as the Fedwire system in the United States, is not subject to systemic risk, its participants are subject to credit risk, especially risk generated by intraday credit exposure. Moreover, liquidity crises are avoided by sharing the debit balance of the failed institutions among the solvent members of the system or by the central bank funding the debit balance. Concerns about intraday credit exposure led the U.S. Federal Reserve to introduce caps on debit positions with Fedwire and CHIPS (Table 2), and to propose interest charges on such debit positions. The presence of a cap on the debit position that an individual bank is allowed to run with Fedwire effectively limits the loss that could be incurred by the Federal Reserve as a result of payment instructions sent out over the Fedwire by a failing bank. However, in a situation where investors have lost confidence in a large money center bank and fail to renew short-term funds (such as maturing certificates of deposit and repurchase agreements), the bank would quickly reach its net debit limit and might then be unable to repay its short-term creditors. As a result, the central bank could be faced with the need to provide funds to the bank through the discount window and hence be once again subject to the credit risk inherent in bank assets being used as collateral.18 Thus, if some banks are regarded as too large to fail, it may be difficult for the central bank to avoid credit risk completely in a liberalized financial system.

Table 2.

Caps on Daylight Overdrafts Across Payment Systems

(Multiples of adjusted primary capital)

article image
Note: Adjusted primary capital for U.S.-chartered banks is the sum of primary capital less all intangible assets and deferred net losses on loans and other assets sold. Source: Federal Reserve Bulletin (November 1987), p. 843.

In contrast, the main international payments system—CHIPS—has only recently adopted payment finality and its members were therefore significantly exposed to the credit risk of lending banks. The 150 member banks of CHIPS are international banks with varying credit ratings. Credit risk arises when banks send out payments for a customer during the day before receiving good funds in the customer’s account. For example, a bank might receive a message from the clearinghouse that an account will be credited with a given amount of dollars at the end of the day. The bank might then be asked by the account holder to make payments to other banks from the account through CHIPS even though the bank has not received good funds. Competition has generally forced banks to be prepared to make such payments. All such payment messages are netted at the end of the day by CHIPS, and net balances are cleared through Fedwire transfers among the settlement banks.19 Thus a central element of the international payments system is the extension of credit among the banks that are members of CHIPS.20

If a disturbance occurs in the financial markets, such as the bankruptcy of a major nonfinancial company, some CHIPS members might be unable to settle their debit balances by borrowing in the interbank market for federal funds. In this case, payments to and from that participant would be unwound, and new net positions would be calculated for the remaining participants. If one of these remaining participants was unable to settle them, this process of calculating new net positions would continue until settlement was achieved. Participants in CHIPS permit most of their customers to use credits for CHIPS payments during the day prior to settlement while reserving the right to charge back such credits if the transferring bank does not settle its CHIPS position. Simulations of the unwinding of transactions under the assumption that one large CHIPS participant would be unable to meet its payments obligations suggest that such failures could drastically change the net positions of other participants, thus inducing a series of failures to settle by the remaining participants.21 Current CHIPS rules and the practice of unwinding could thus potentially contribute to systemic risks in the banking system and put pressure on the Federal Reserve to provide liquidity assistance while losses and solvency problems are determined.

Since the Federal Reserve does not regulate and supervise the foreign members of CHIPS, it could only guarantee the domestic transactors on CHIPS (as is done on Fedwire). Moreover, attempts by CHIPS itself to impose regulations on its foreign member banks would require the approval of bank regulators in those countries. Under current arrangements, the failure of a major international banking institution could nonetheless cause a systemic crisis if it were to spread illiquidity across the CHIPS system.

Policy Initiatives to Reduce Risk in Payment Systems

This section reviews the major policy initiatives undertaken to reduce credit risk arising from daylight overdrafts in gross payment systems and to reduce systemic risk in net payment systems. The thrust of these initiatives has been to reduce the credit exposure of the clearinghouse, such as Fedwire, and to introduce payments finality in net settlement systems, such as CHIPS, as well as to strengthen the operational, financial, and liquidity characteristics of the private net settlement systems.

In 1986, the U.S. Federal Reserve Board introduced a risk reduction program for the payments system that focused on controlling the direct Federal Reserve credit risk exposure arising with the extension of intraday credit on Fedwire (see Tables 2 and 3). The policy was later extended to CHIPS, in that it established a maximum amount of intraday overdraft that depository institutions are permitted to incur over Fedwire and private large dollar payment systems such as CHIPS. Such caps are defined as a multiple of the depository institution’s adjusted primary capital and are based on the institution’s self-evaluation of its creditworthiness, credit policies, and operational controls. This policy was further strengthened in July 1987 when the Board adopted a two-step 25 percent reduction in cross-system net debit caps. In addition to the cross-system caps, which apply to the daylight overdraft position on Fedwire and CHIPS, caps apply to the sum of a bank’s overdraft on its reserve account and its net debit position on CHIPS at each moment during the day. Each bank places itself in one of four self-assessment categories to determine its caps for both a one-day and a two-week average maximum daylight overdraft as a percentage of primary adjusted capital (see Table 2).

Table 3.

Summary of U.S. Federal Reserve Board Proposals for Risk Reduction in Payments System

article image
Source: United States, Board of Governors of the Federal Reserve System (1989).

In addition, the Federal Reserve requires each participant on CHIPS to set a limit on its net credit position vis-à-vis each of the other participants in the system. Finally, CHIPS is required to establish limits on the net debit positions of each participant with all other participants in the system. This limit is set at 5 percent of the sum of all bilateral credit limits for a given participant extended by all other CHIPS participants.22

In June 1988, the Board of Governors Large-Dollar Payments System Advisory Group made recommendations on how to achieve further reductions in payment system risks. It concluded that (1) some level of public and private intraday credit was desirable to absorb the inevitable lack of synchronization of payments flows; (2) the unconstrained access to Federal Reserve overdrafts induces private institutions to use more intraday credit than is optimal; (3) either caps or pricing can in principle be used to discourage the use of daylight overdrafts; and (4) binding caps will produce a more volatile intraday rate than pricing. With regard to the pricing of daylight overdrafts, it has been proposed that the Board charge 25 basis points on daily average overdrafts, less a deductible equal to 10 percent of capital. This fee would be phased in over a three-year period beginning in 1991. The Large-Dollar Payments System Advisory Group judged that this price should be set so as to “induce the creation of a private-sector market to replace much of Federal Reserve funding of intraday credit.”23 It is expected that the assessment of a fee for daylight overdrafts will create incentives for depository institutions to reduce their demand for intraday extensions of credit for transfers of funds by reserve banks, thus reducing the reserve banks’ direct credit exposure.24 Such a reduction in daylight overdrafts could be achieved by delaying less critical payments, shifts of payments to CHIPS, greater use of netting, and the development of an intraday federal funds market.

In view of the anticipated shift of settlement from Fedwire to CHIPS, the Federal Reserve has also been concerned with reducing the systemic risks associated with private sector large dollar payment and clearing systems. It has therefore encouraged the New York Clearing House to adopt settlement finality for CHIPS. If fully adopted, settlement finality assures that CHIPS net debits and credits will be settled each day even if some participants are unable to settle. This procedure therefore separates liquidity from credit risk, since the allocation of losses can be resolved later. In consultation with the Board of Governors of the Federal Reserve System, the New York Clearing House recommended that all CHIPS participants sign a loss-sharing agreement to provide funds necessary to complete CHIPS settlements if one or a limited number of participants are unable to meet their respective debit balances. In addition, if any participant fails, or is otherwise unable to meet his balance, the remaining participants would make up the shortfall. They will contribute amounts computed by a formula under which each bank makes up a portion of the failed participant’s balance in relation to its own credit judgment regarding the failed banks, that is, the bilateral limit, which enables the participants to incur the obligation on the system. This procedure is also expected to improve the assessment of bilateral credit limits.

An important element involved in instituting payment finality is that the multilateral netting rules on CHIPS be upheld legally. CHIPS and its individual participants cannot take the risk that the general creditors of a failed participant bank will demand immediate payment of the gross amount of payment sent to the failed bank, while treating payments made by the failed bank as unsecured (and possibly depositor-subordinated) claims against the receivership.

Finality of payment rules, such as those proposed for CHIPS, internalize the costs of a settlement failure and thus provide incentives for market participants to control payment risks. Such finality rules are of particular interest now because the U.S. National Conference of Commissioners on Uniform State Laws is currently drafting provisions of the uniform commercial code that codify a law of electronic funds transfer (EFT). This effort will largely determine the future statutory environment that guides the rights, obligations, and risk assignments of network participants. Finality rules need to specify with certainty who pays the cost of a settlement failure.

In the absence of established law and precedent, the New York Clearing House is seeking rulings concerning the proposed arrangements from the Federal Reserve and other regulatory agencies, and has proposed an amendment to the Federal Reserve Act. It is envisioned that settlement finality will become fully operative in 1991. While settlement finality on CHIPS will reduce private credit risk, it does not entirely eliminate it. It is possible, for example, that the debit balance of a group of failed banks is larger than that covered by the contributions of the remaining CHIPS participants.

One further possibility for eliminating the direct credit exposure of the central bank is to establish payment queues with payment transfers being held in the queue until covering funds have arrived, rather than allowing overdrafts. Hence, a payment without temporary cover is not rejected but is instead placed in a queue and then processed on a first-in first-out basis after covering funds are received. However, as noted above, the problem with adopting such a centralized payment queuing system is that, without reserve requirements or some price incentive, there is no guarantee that the level of reserves held voluntarily will remain above that required to avoid grid lock of the system.

An important public policy issue in reducing payments risk relates to the growth of offshore clearance and settlement systems. This issue has led to the preparation of a report on such systems by a group of experts on payment systems of the central banks of the Group of Ten countries.25 The incentive to develop offshore payment networks or netting systems derives from a less stringent regulatory environment and from significant time zone differences between North America, Asia, and Europe. Offshore dollar arrangements must ultimately settle in the United States through either CHIPS or Fedwire since good funds can only be transferred in accounts held with the U.S. Federal Reserve Banks. Significant disruptions in offshore clearance and settlements systems for foreign exchange and securities owing to the failure of a participating institution could well result in systemic problems in the United States and other major countries. Offshore clearing of U.S. dollar payments for subsequent net settlement in the United States has been viewed as obscuring and possibly increasing the level of systemic risk in the U.S. large dollar payments system and in the international settlements process.

Finally, offshore multilateral netting arrangements complicate the allocation of supervisory responsibilities. Formalized netting arrangements and offshore payments systems (that is, groupings of individual banks with interrelated credit and liquidity risks) have shifted risks among participants, and it is at times unclear at which level a supervisor should examine the credit, liquidity, and operational risks facing these institutions. Both the host country authorities of an offshore system and the home country of the multinational participants in that system have an interest in supervising it if it affects the solvency and liquidity of their institutions. In addition, the central bank for the currency that is being cleared in the offshore system could have a supervisory interest in the system for monetary policy reasons.

Another concern is that a shift away from the use of the central payments system toward the offshore netting system might amount to the decentralization of the major monetary mechanisms and thus undermine the relationship between key monetary aggregates and domestic activity. In particular, netting arrangements can serve as a close substitute for money in terms of economizing on transaction costs.26

In response to these concerns, the U.S. Federal Reserve Board published a draft interim policy statement on offshore dollar clearing and netting systems.27 This statememt indicated that (1) any subsystem that settled directly or indirectly through CHIPS or Fedwire should be subject to oversight as a system by a relevant central bank or supervisory authority; (2) participants should clearly identify the operational, liquidity, and credit risks created within the system; (3) the system should provide for the finality of settlement obligations; and (4) the direct or indirect settlement of the system’s obligations through CHIPS or Fedwire should be done by a U.S. settlement bank.

The increasing globalization of securities markets and foreign exchange markets and the shift toward 24-hour trading in certain types of securities have also led to proposals for extending settlement hours, preferably toward a 24-hour settlement period. Such an option is currently under review by the Federal Reserve System. For example, in foreign exchange transactions, the delivery of each currency is made at a bank in the country that issues the currency. If the banking hours of the respective central banks do not overlap, a time lag will appear between final settlements in the home currencies. Currently settlement of spot transactions in foreign exchange markets occurs on the second day after the contract has been entered into.28

The counterparty risks in the settlement process for foreign exchange transactions could be minimized by moving toward a system based on delivery against payment. However, if a yen-U.S. dollar foreign exchange transaction took place in New York, such a simultaneous matching of the yen settlement to the dollar settlement would require the Bank of Japan to effect the interbank settlement during New York daytime or during nighttime in Japan. Some major U.S. correspondent clearing bank would then have to make transfers into a dollar account during the night.

Without 24-hour settlement, offshore settlement systems have developed, with the attendant liquidity and credit risks. The inability to make a Fedwire transfer outside the 9:00 a.m. to 6:30 p.m. eastern time slot has led banks to use other networks to make dollar payments.29

Summary and Conclusions

Efficient and stable payment systems are of fundamental importance in maintaining an orderly international monetary system. Major disruptions of national and international payment systems would have highly adverse effects on international trade, capital flows, and real activity. A key issue now being addressed by the authorities in a number of major countries is whether existing institutional arrangements need to be modified to manage better the sharp increases in the volume of payments transactions and the liquidity and credit risks that have arisen from the expansion of international trade and capital flows and the growing integration of major financial markets. For example, the direct credit exposure in the form of daylight overdrafts on the world’s largest payments system, the Fedwire, has grown significantly during the last decade, as daily average overdrafts have risen to about $65 billion.

The internationalization of financial markets and of clearing arrangements has further contributed to systemic risk. Clearinghouses directly link the financial solvency of members through credit and debit positions in the clearinghouse. Periodic settlements are major points in time where the liquidity and solvency of banks are tested by the market. Many major international banks play the role of settlement banks for international transactions undertaken by smaller regional or local national banks. Thus, disturbances in one payments system would tend to spread rapidly to other systems.

One source of concern has been the growth of international or offshore netting arrangements, in particular in foreign exchange markets. Netting arrangements that are located outside the country whose currency is being netted have the potential to alter significantly the structure of the international interbank clearing and settlement process. Multilateral netting schemes without settlement finality in which participants retain some responsibility for the gross transactions are particularly exposed to credit risks. The unwinding necessary if a participant is unable to settle is a significant source of systemic risk. Hence, netting by novation is emphasized, which reduces the overall credit exposure of participants to their net position vis-à-vis the clearinghouse. Furthermore, novation with settlement finality is the solution desired by regulatory authorities.

Since many of the credit and liquidity risks cannot be directly controlled by individual participants, it is unlikely that market forces alone would produce international payment arrangements that would adequately manage the systemic risks. Private cooperative arrangements without central bank involvement are unlikely to reduce systemic risk to acceptable levels, because the power of private clearinghouses to impose restrictions on members, as well as to provide the liquidity occasionally required by members at closing time, is limited. Furthermore, the ability of private financial institutions to undertake regulatory arbitrage (to relocate activities to a less regulated environment) suggests that cooperation between major central banks will have to be an important element in the management of payments system risks. Central bank cooperation in strengthening international payments would complement the cooperation already achieved in the area of bank supervision and monetary policy.30 Nonetheless, the allocation of supervisory responsibility among the various national supervisory authorities remains unresolved. In part, this reflects the fact that while the efficiencies of netting arrangements may be enjoyed by banks located in one country, the credit and liquidity risks associated with the settlement of payments resulting from that netting system may be experienced in the banking system of another country. This issue is of particular concern to U.S. authorities, as most of the offshore netting arrangements are subsequently settled through CHIPS and Fedwire.

These developments have led monetary authorities, particularly the U.S. Federal Reserve Board, to design policy initiatives to reduce risks in payment systems. The Federal Reserve’s program of risk reduction in the payments system began in 1986 and has focused on controlling levels of intraday credit extension on both U.S. large dollar payment systems, Fedwire and CHIPS, by limiting the total net debit position of participants in the systems. In addition, the Federal Reserve proposed that intraday credit on Fedwire be subject to a 25-basis points interest charge, and CHIPS began to introduce settlement finality in late 1990.

Since the proposed constraints on Fedwire and private payment systems in the United States are likely to stimulate the use of other private netting systems and offshore networks, the Central Bank Committee of the Group of Ten produced a series of proposals to strengthen such netting schemes, with emphasis on the finality of payments. Furthermore, the globalization of financial markets and the development of continuous 24-hour trading in some spot and futures contracts may necessitate operating the major clearing systems such as Fedwire on a 24-hour basis to avoid the systemic risk associated with current and proposed offshore large dollar payment networks.

A key element containing and reducing systemic risk in payment systems is the finality of payments—reducing the period during which balances remain unsettled. In addition, improvements in the operational efficiency and financial resources of clearing and settlement systems are important in containing credit and liquidity risks. Efficient operational features allow participants to keep track of exposures, including intraday positions. A clearing entity should have in place the financial resources and commitments (in the form of reserves, collateral, committed bank loans, guarantees, etc.) to ensure that if a participant defaults, settlement will still take place—that is, no unwinding of the transaction will occur. The financial resources of payment systems are reflected in the capital structure of member firms, the level of margins, and access to liquidity and collateral. A strengthening of these features is deemed to increase the ability of the institutions to withstand disturbances and thereby reduce payments system risk.

Appendix I

Table 4.

United States: Payments System, 1987

article image
Note: EFT = electronic funds transfers; ACH = automatic clearinghouses; ATM = automated teller machines. Source: Bank for International Settlements (1989b).

Includes traveler’s checks (1.4 billion with a value of $47 billion) and money orders (0.8 billion valued at $70 billion).

The Nilson Report, May 1988 (Los Angeles, California). Includes all types of credit card transactions; bank card volume: 2.5 billion valued at $165.3 billion. Credit card payment volume and value included in check data.

Includes Fedwire volume of 53 million valued at $142 trillion and CHIPS volume of 31 million valued at $139 trillion.

Approximately 40 percent of the dollar value of Fedwire transfers are for interbank loan transactions, 10 percent for Eurodollar transactions, and 10 percent for commercial transactions, whereas 55 percent of the dollar value of CHIPS transactions are for foreign exchange transactions and 28 percent for Eurodollar transactions.

ACH credit payments: 584 million transactions with a value of $803 billion; ATM payments: 29 million transactions with a value of $2 billion.

Table 5.

Japan: Payments System, 19871

article image
Source: Bank for International Settlements (1989b).

Estimated figures.

Table 6.

Germany: Payments System, 1987

article image
Note: EFT = electronic funds transfers; ATM = automated teller machines. Source: Bank for International Settlements (1989b).

Partly estimated.

Not included in direct debits to avoid double counting.

Charge cards and bank cards, excluding retail cards; the card companies’ settlements with the retailers (normally credit transfers) and payment of the monthly totals by cardholders to card issuers by credit transfers, direct debit, or check are contained in the corresponding items.

Excluding interbank transfers. Interbank transfers via Bundesbank, partly estimated:

article image

Including customers’ paper-based credit transfers that were routed into the paperless procedure by the bank to which they were first submitted.

Including cash dispenser/ATM withdrawals made with EC-cards at banks other than the one issuing the card.

Table 7.

United Kingdom: Payments System, Year-End 1987

article image
Source: Bank for International Settlements (1989b).

Excluding an estimated 300 million cashed checks, valued at $25 billion (some £1 3 billion).

Including interbranch checks.

Excluding transactions by holders of an estimated 9 million charge and budget cards issued by retailers, but including transactions by holders of over 1.5 million travel and entertainment cards.

Including standing orders.

Via Clearing House Automated Payment System (CHAPS).

Excluding government payments in cash from post offices against state benefit vouchers.

Table 8.

France: Payments System, 19871

article image
Note: EFT = electronic funds transfers. Source: Bank for International Settlements (1989b).

The figures in this table combine the data relating to all payments instruments, whether they are routed via “official” circuits or not.

Including postal checks.

Of which 45 percent (by volume) did not give rise to electronic payment.

These figures include credit transfers of a purely interbank nature that could not be isolated.

A breakdown is not available.

Including the universal payment order.

Table 9.

Switzerland: Payments System, 1987

article image
Note: EFT = electronic funds transfers. Source: Bank for International Settlements (1989b).

Euro, bank, Swiss Bankers Traveller, and postal checks.

Rough estimates.

Without payments by debit cards.

Total giro transfers including interbank payments.

In millions of Swiss francs.

Appendix II Example of Systemic Risk

Consider three participating banks (A, B, and C) on the CHIPS system. These banks receive and send payments during the day as follows:

article image

Assume that A issues an order to pay $5 to B; B makes a $10 payment to C; and C makes a payment of $10 to A. Then C makes an additional payment of $15 to A. Finally, A makes a payment of $10 to B; all these payments are made during the course of one settlement day. After CHIPS clears these payments: A is a net creditor for $10, B is a net creditor for $5, and C has a net debt balance of $15. Thus C would have to send $15 over Fedwire into the CHIPS clearing account at the Federal Reserve either directly or through its clearing bank. CHIPS would then pay out $10 to A’s account and $5 to B’s account. However, if C was unable to pay its settlement obligation, CHIPS would unwind the day’s transactions and delete B’s receipts and payments from the relevant net payments. The new transaction structure would become:

article image

Thus A would be left with a $15 net debit position as opposed to a $10 net credit position. If A does not have $15 in its Federal Reserve account, it has to borrow from the federal funds market. However, the bankruptcy of a major bank might make it difficult for a net debit bank to borrow, particularly if its net debit is large relative to its capital. Simulation exercises31 have shown that the unwinding of transactions of one large CHIPS participant would make a high percentage of other participants unable to meet their commitments on CHIPS without acquiring additional reserves in the federal funds market. In such a situation, the Federal Reserve could lend reserves to bank A against collateral, or, at an earlier stage, to bank C so that the initial settlement could go through.

References

  • Anvari, Mohsen, “The Canadian Payment System: An Evolving Structure,” in The U.S. Payment System: Efficiency, Risk and the Role of the Federal Reserve, ed. by David B. Humphrey (Boston: Kluwer, 1990).

    • Search Google Scholar
    • Export Citation
  • Association of Reserve City Bankers (ARCB), Finality in Private Payments Systems: Issues and Alternatives (Washington: ARCB, December 1985).

    • Search Google Scholar
    • Export Citation
  • Association of Reserve City Bankers (ARCB), Report of the Working Group of the Association of Reserve City Bankers on Book-Entry Daylight Overdrafts (Washington: ARCB, June 1986).

    • Search Google Scholar
    • Export Citation
  • Banca d’Italia, White Paper on the Payment System in Italy (Rome: Banca d’Italia, April 1988).

  • Bank for International Settlements (1989a), Report on Netting Schemes, Prepared by the Group of Experts on Payment Systems of the central banks of the Group of Ten countries (Basle, February 1989).

    • Search Google Scholar
    • Export Citation
  • Bank for International Settlements (1989b), Payment Systems in Eleven Developed Countries (Basle, 3rd ed., May 1989).

  • Bank of Canada, “Matters Arising from Transactions of the Northland Bank with its Clearing Agent The Royal Bank of Canada, and with the Bank of Canada on August 30th, 1985” (Press Statement, September 18, 1985).

    • Search Google Scholar
    • Export Citation
  • Bank of Japan, “Japanese Transfer Systems in the Era of Financial Deregulation and Globalization” (July 1989).

  • Baxter, William F., “Bank Interchange of Transactional Paper: Legal and Economic Perspectives,” Journal of Law and Economics, Vol. 26 (October 1983).

    • Search Google Scholar
    • Export Citation
  • Belton, Terrence M., and others, “Daylight Overdrafts and Payments System Risk,” Federal Reserve Bulletin, Vol. 73 (November 1987).

  • Benston, George J., and George G. Kaufman, “Regulating Bank Safety and Performance,” in Restructuring Banking and Financial Services in America, ed. by William S. Haraf and Rose Marie Kushmeider (Washington: American Enterprise Institute, 1988).

    • Search Google Scholar
    • Export Citation
  • Benston, George J., and George G. Kaufman, Robert Eisenbeis, Paul Horvitz, Edward Kane, and George Kaufman, Perspectives on Safe and Sound Banking: Past, Present, and Future (Cambridge, Massachusetts: MIT Press, 1986).

    • Search Google Scholar
    • Export Citation
  • Brickley, James A., and Christopher M. James, “Access to Deposit Insurance, Insolvency Rules and the Stock Returns of Financial Institutions,” Journal of Financial Economics, Vol. 16 (July 1986).

    • Search Google Scholar
    • Export Citation
  • Brown, John Prather, “Toward an Economic Theory of Liability,” Journal of Legal Studies, Vol. 2 (June 1973).

  • Cannon, James G., Clearing Houses: Their History, Models, and Administration (London: Smith, Elder, and Company, 1901), reprinted in National Monetary Commission, Vol. 6 (Washington, 1911).

    • Search Google Scholar
    • Export Citation
  • Clarke, John J., “An Item is an Item is an Item: Article 4 of the U.C.C. and the Electronic Age,” Business Lawyer, Vol. 25 (November 1969).

    • Search Google Scholar
    • Export Citation
  • Cooter, Robert D., and Edward L. Rubin, “A Theory of Loss Allocation for Consumer Payments,” Texas Law Review, Vol. 66 (November 1987).

    • Search Google Scholar
    • Export Citation
  • Corrigan, E. Gerald, “Financial Market Structure: A Longer View,” Federal Reserve Bank of New York, Seventy-Second Annual Report For the Year Ended December 31, 1986 (1987).

    • Search Google Scholar
    • Export Citation
  • de Roover, Raymond, The Dawn of Modern Banking (New Haven, Connecticut: Yale University Press, 1979).

  • Dudley, William C, “Controlling Risk on Large-Dollar Wire Transfer Systems,” in Technology and the Regulation of Financial Markets: Securities, Futures, and Banking, ed. by Anthony Saunders and Lawrence J. White (Lexington, Massachusetts: Lexington Books, 1986).

    • Search Google Scholar
    • Export Citation
  • Edwards, Franklin R., “The Future Financial Structure: Fears and Policies,” in Restructuring Banking and Financial Services in America, ed. by William S. Haraf and Rose Marie Kushmeider (Washington: American Enterprise Institute, 1988).

    • Search Google Scholar
    • Export Citation
  • Eisenbeis, Robert A., “Eroding Market Imperfections: Implications for Financial Intermediaries, the Payments System, and Regulatory Reform,” in Restructuring the Financial System (Federal Reserve Bank of Kansas City, 1987).

    • Search Google Scholar
    • Export Citation
  • Fama, Eugene, “Banking in a Theory of Finance,” Journal of Monetary Economics, Vol. 6 (January 1980).

  • Fama, Eugene, “What’s Different About Banks?” Journal of Monetary Economics, Vol. 15 (January 1985).

  • Faulhaber, Gerald R., Almarin Phillips, and Anthony M. Santomero, “Payment Risk, Network Risk, and the Role of the Fed,” in The U.S. Payment System: Efficiency, Risk and the Role of the Federal Reserve, ed. by David B. Humphrey (Boston: Kluwer, 1990).

    • Search Google Scholar
    • Export Citation
  • Federal Reserve Bank of New York, “Large-Dollar Payment Flows from New York,” Federal Reserve Bank of New York, Quarterly Review, Vol. 12 (Winter 1987–88).

    • Search Google Scholar
    • Export Citation
  • Federal Reserve Bank of New York, Clearing and Settling the Euro-Securities Market: Euro-Clear and Cedel (New York, March 1989).

  • Federal Reserve Bank of New York, An Overview of the Operation of the Options Clearing Corporation (New York, April 1989).

  • Federation Internationale des Bourses de Valeurs, Improving International Settlement, Report of the Task Force appointed by the FIBV (1989).

    • Search Google Scholar
    • Export Citation
  • Flannery, Mark J., “Payments System Risk and Public Policy,” in Restructuring Banking and Financial Services in America, ed. by William S. Haraf and Rose Marie Kushmeider (Washington: American Enterprise Institute, 1988).

    • Search Google Scholar
    • Export Citation
  • Folkerts-Landau, D., “Coordination of Financial Policy,” in International Policy Coordination and Exchange Rate Fluctuations, ed. by W. Branson, J.A. Frenkel, and M. Goldstein (Chicago: University of Chicago Press, 1990).

    • Search Google Scholar
    • Export Citation
  • Friedman, Milton, A Program for Monetary Stability (New York: Fordham University Press, 1959).

  • Friedman, Milton, and Anna J. Schwartz, A Monetary History of the United States, 1867–1960 (Princeton, New Jersey: Princeton University Press, 1963).

    • Search Google Scholar
    • Export Citation
  • Garbade, Kenneth D., and William L. Silber, “The Payment System and Domestic Exchange Rates: Technological Versus Institutional Change,” Journal of Monetary Economics, Vol. 5 (January 1979).

    • Search Google Scholar
    • Export Citation
  • Gelfand, Mathew D., and David E. Lindsey, “The Simple Microanalytics of Payments System Risk,” Finance and Economics Discussion Series, No. 61 (Washington: Federal Reserve Board, March 1989).

    • Search Google Scholar
    • Export Citation
  • Goodfriend, Marvin S., “Central Banking Under the Gold Standard,” in Money Cycles, and Exchange Rates: Essays in Honor of Allan H. Meltzer, ed. by Karl Brunner and Bennett McCallum, Carnegie-Rochester Conference Series on Public Policy, Vol. 29 (1988).

    • Search Google Scholar
    • Export Citation
  • Goodfriend, Marvin S., “Money, Credit, Banking, and Payment System Policy,” in The U.S. Payment System: Efficiency, Risk and the Role of the Federal Reserve, ed. by David B. Humphrey (Boston: Kluwer, 1990).

    • Search Google Scholar
    • Export Citation
  • Goodfriend, Marvin S., and Robert G. King, “Financial Deregulation, Monetary Policy, and Central Banking,” in Restructuring Banking and Financial Services in America, ed. by William S. Haraf and Rose Marie Kushmeider (Washington: American Enterprise Institute, 1988).

    • Search Google Scholar
    • Export Citation
  • Goodhart, C.A.E., “Why Do Banks Need a Central Bank?” Oxford Economic Papers, Vol. 39 (March 1987).

  • Goodhart, C.A.E., The Evolution of Central Banks, (Cambridge, Massachusetts: MIT Press, 1988).

  • Gorton, Gary, and Donald J. Mullineaux, “The Joint Production of Confidence: Endogenous Regulation and Nineteenth Century Commercial-Bank Clearinghouses,” Journal of Money, Credit, and Banking, Vol. 19 (November 1987).

    • Search Google Scholar
    • Export Citation
  • Greenspan, Alan, “International Payment Systems Developments,” speech at the International Symposium on Banking and Payment Services (Washington, June 9, 1989).

    • Search Google Scholar
    • Export Citation
  • Group of Thirty, Clearance and Settlement Systems in the World’s Securities Markets (New York: Group of Thirty, 1989).

  • Guttentag, Jack M., and Richard J. Herring, Disaster Myopia in International Banking, Brookings Discussion Papers in International Economics, No. 31 (Washington: Brookings Institution, 1985).

    • Search Google Scholar
    • Export Citation
  • Heurtas, Thomas F., “Risk in the Payments System,” in U.S. Congress, House, Committee on Government Operations, Structure and Regulation of Financial Firms and Holding Companies (Part 3), 99th Congress, 2nd Session (Washington: Government Printing Office, 1987), pp. 36189.

    • Search Google Scholar
    • Export Citation
  • Humphrey, David B., The U.S. Payments System: Costs, Pricing, Competition, and Risk, Monograph Series in Finance and Economics (New York: New York University, 1984).

    • Search Google Scholar
    • Export Citation
  • Humphrey, David B., “Payments Finality and Risk of Settlement Failure,” in Technology and the Regulation of Financial Markets: Securities, Futures, and Banking, ed. by Anthony Saunders and Lawrence J. White (Lexington, Massachusetts: Lexington Books, 1986).

    • Search Google Scholar
    • Export Citation
  • Ireland, Oliver I., “Payment System Risk,” in U.S. Congress, House, Committee on Government Operations, Structure and Regulation of Financial Firms and Holding Companies (Part 1), 99th Congress, 2nd Session (Washington: Government Printing Office, 1987), pp. 500508.

    • Search Google Scholar
    • Export Citation
  • James, Christopher, “Some Evidence on the Uniqueness of Bank Loans,” Journal of Financial Economics, Vol. 19 (December 1987).

  • Jensen, Michael C., and William H. Meckling, “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure,” Journal of Financial Economics, Vol. 3 (October 1976).

    • Search Google Scholar
    • Export Citation
  • Jevons, William S., Money and the Mechanism of Exchange (New York: Appleton, 1898).

  • Johnson, Manuel H., “Challenge to the Federal Reserve in the Payments Mechanism,” Issues in Bank Regulation (Summer 1988).

  • Kantrow, Yvette D., “Big NY Banks May Spin Off Chips Network,” American Banker, Vol. 23 (June 27, 1988).

  • Kareken, John H., “Ensuring Financial Stability,” in The Search for Financial Stability: The Past Fifty Years (San Francisco: Federal Reserve Bank of San Francisco, 1985).

    • Search Google Scholar
    • Export Citation
  • Kessler, J., Study on Improvements in the Settlement of Cross-Border Securities Transactions in the EEC (Commission of the European Communities, July 1988).

    • Search Google Scholar
    • Export Citation
  • King, Robert G., and Charles I. Plosser, “Money as the Mechanism of Exchange,” Journal of Monetary Economics, Vol. 17 (January 1986).

    • Search Google Scholar
    • Export Citation
  • Lindow, Wesley, “Bank Capital and Risk Assets,” National Banking Review, Comptroller of the Currency, U.S. Treasury Department (September 1963).

    • Search Google Scholar
    • Export Citation
  • Litan, Robert E., “Taking the Dangers Out of Bank Deregulation,” Brookings Review, Vol. 4 (Fall 1986).

  • Lumpkin, Stephen A., “Repurchase and Reverse Repurchase Agreements,” in Instruments of the Money Market, ed. by Timothy Q. Cook and Timothy D. Rowe (Richmond, Virginia: Federal Reserve Bank of Richmond, 6th ed., 1986).

    • Search Google Scholar
    • Export Citation
  • McCallum, Bennett, “Bank Deregulation, Accounting Systems and Exchange, and the Unit of Account: A Critical Review,” in The New Monetary Economics, Fiscal Issues and Unemployment, ed. by Karl Brunner and Allan H. Meltzer, Carnegie-Rochester Conference Series on Public Policy, Vol. 23 (1985).

    • Search Google Scholar
    • Export Citation
  • Mengle, David L., “Legal and Regulatory Reform in Electronic Payments: An Evaluation of Payment Finality Rules,” in The U.S. Payment System: Efficiency, Risk and the Role of the Federal Reserve, ed. by David B. Humphrey (Boston: Kluwer, 1990).

    • Search Google Scholar
    • Export Citation
  • Mengle, David L., David B. Humphrey, and Bruce J. Summers, “Intraday Credit: Risk, Value, and Pricing,” Federal Reserve Bank of Richmond Economic Review, Vol. 73 (January/February 1987).

    • Search Google Scholar
    • Export Citation
  • Mullineaux, Donald J., “Competitive Monies and the Suffolk Bank System: A Contractual Perspective,” Southern Economic Journal, Vol. 53 (April 1987).

    • Search Google Scholar
    • Export Citation
  • New York Clearing House Association, Constitution of the New York Clearing House Association (May 1903).

  • New York Clearing House Association, Memorandum on “Increased Assurances of Settlement Formality or CHIPS” (April 1989).

  • Padoa-Schioppa, T., “Payment System: A New Ground for Central Bank Cooperation,” unpublished (June 1989).

  • “Proposals for International Convergence of Capital Measurement and Capital Standards,” Issues in Bank Regulation (Winter 1988).

  • Rothschild, Michael, and Joseph E. Stiglitz, “Increasing Risk: I. A Definition,” Journal of Economic Theory, Vol. 2 (September 1970).

    • Search Google Scholar
    • Export Citation
  • Schwartz, Anna J., “Financial Stability and the Federal Safety Net,” in Restructuring Banking and Financial Services in America, ed. by William S. Haraf and Rose Marie Kushmeider (Washington: American Enterprise Institute, 1988).

    • Search Google Scholar
    • Export Citation
  • Timberlake, Richard H., Jr., “The Central Banking Role of Clearinghouse Associations,” Journal of Money, Credit and Banking, Vol. 16 (February 1984).

    • Search Google Scholar
    • Export Citation
  • United States, Board of Governors of the Federal Reserve System, “Policy Statement Regarding Risks on Large-Dollar Wire Transfer Systems,” Federal Register, Vol. 50, No. 99 (May 22, 1985).

    • Search Google Scholar
    • Export Citation
  • United States, “Interim Policy Statement Regarding Risks on Large-Dollar Wire Transfer Systems,” Federal Register, Vol. 52, No. 151 (August 6, 1987).

    • Search Google Scholar
    • Export Citation
  • United States (1988a), Large-Dollar Payments System Advisory Group, A Strategic Plan for Managing Risk in the Payments System, Report of the Large-Dollar Payments System Advisory Group to the Payments System Policy Committee of the Federal Reserve System (Washington, August 1988).

    • Search Google Scholar
    • Export Citation
  • United States (1988b), Controlling Risk in the Payments System, Report of the Task Force on Controlling Payments System Risk to the Payments System Policy Committee of the Federal Reserve System (Washington, August 1988).

    • Search Google Scholar
    • Export Citation
  • United States, Proposals for Modifying the Payments System Risk Reduction Policy of the Federal Reserve System, Prepared by the Staff of the Federal Reserve System (Washington, May 1989).

    • Search Google Scholar
    • Export Citation
  • United States, Congress, House, Committee on Government Operations and Committee on Banking, Finance, and Urban Affairs, The Role of the Federal Reserve in Check Clearing and the Nation’s Payments System, Joint Hearings before Subcommittees of the Government Operations and Banking, Finance and Urban Affairs Committees, 98th Congress, 1st Session, June 15, 1983 (Washington: Government Printing Office, 1983).

    • Search Google Scholar
    • Export Citation
  • Van Hoose, David, “The Angell Proposal: An Overview” (Board of Governors of the Federal Reserve System, June 6, 1988).

  • Vital, Christian, and David L. Mengle, “SIC: Switzerland’s New Electronic Interbank Payment System,” Federal Reserve Bank of Richmond Economic Review (November/December 1988).

    • Search Google Scholar
    • Export Citation
  • Westerfield, Ray B., Banking Principles and Practice, Vol. 3 (New York: Ronald Press Company, 1921).

  • Whitney, D.R., The Suffolk Bank (Cambridge, Massachusetts: Riverside Press, 1878).

  • Williamson, Stephen D., “Bank Failures, Financial Restrictions, and Aggregate Fluctuations: Canada and the United States, 1870–1913,” Federal Reserve Bank of Minneapolis Quarterly Review (Summer 1989).

    • Search Google Scholar
    • Export Citation
  • Young, John E., “The Rise and Fall of Federal Reserve Float,” Federal Reserve Bank of Kansas City Economic Review, Vol. 71 (February 1986).

    • Search Google Scholar
    • Export Citation
1

The evolution of modern banking systems received strong impetus from the demand for an efficient mechanism to facilitate the payments flows needed to sustain a growing economy. The transfer of securities among buyers and sellers in spot and futures markets has also received increased attention recently (see Group of Thirty (1989) and Kessler (1988)).

2

Tables 49 of Appendix I summarize the main characteristics of major payment systems in the industrial countries.

3

For example, in the United States, paper checks account for about 95 percent of noncurrency payments but for only 14 percent of dollar volume, whereas electronic payments account for 0.1 percent of the number of transactions, but for 80 percent of the dollar volume.

4

The importance of the wholesale payments system in the United States to the international economy is a reflection of the growth of the liquid and deep money and securities markets where trading activity gives rise to large transfers of funds. Furthermore, the U.S. dollar system is also the major system for transfers of international funds because of the dollar’s role as a vehicle currency and the importance of Eurodollar markets; see also the following section of this study.

6

Timberlake (1984). The New York Clearing House was founded in 1853 by about 50 member banks.

8

Schwartz (1988) has argued that in the United States there were at least 17 banking crises during the period from 1793 to 1933, but none have occurred since 1933, the beginning of active Federal Reserve intervention.

9

For example, Corrigan (1987) argued that “the efficient working of a large modern economy clearly requires the presence of a stock of financial assets which are highly liquid and readily transferable, thereby facilitating the broad range of transactions needed to sustain the real and financial sectors of the economy. To be highly liquid, such assets must be available to the owner at very short notice (a day or less) at par. To be readily transferable, ownership rights in such assets must be capable of being readily shifted to other economic agents, also at par and in a form in which they are acceptable by that other party” (pp. 12–13).

10

Some of these concerns have been discussed in recent conferences and symposia: the Group of Thirty Symposium on Clearance and Settlement Issues in the Global Securities Markets in London in March 1988; the International Symposium on Banking and Payment Services sponsored by the Board of Governors of the U.S. Federal Reserve System in June 1989; and the Williamsburg Payments System Symposium of the Federal Reserve Bank of Richmond in May 1988.

11

The Bank of New York, a major clearing bank in the U.S. payments system, experienced a computer breakdown on November 21, 1985, which led the U.S. Federal Reserve to make an overnight loan of $22.6 billion from the discount window, collateralized by $36 billion in securities.

13

It is possible to refine the concept of finality further into payments, settlement, and receiver finality, depending on the stage of the funds transfer transaction (see United States, Board of Governors of the Federal Reserve System (1989)).

14

Dealers generally make commitments to deliver specific securities by the end of the day. While the customer receives interest on the promised securities for the day, he pays the dealer only when the securities are delivered. Failure to deliver would expose the dealer to losses, and, to minimize the probability of such losses, the dealer waits until early afternoon before directing his clearing bank to send the securities sold to the book entry accounts of the banks of the purchasers. When the investor in a repurchase agreement returns the securities to the dealer, the dealer’s securities account at the clearing bank increases and his demand deposit account decreases, while at the Federal Reserve Bank the book entry account of the clearing bank increases and its reserve account decreases. The dealer will also finance his inventory of securities by overdrawing his account with the clearing bank. The overdraft on the reserve and deposit accounts is reversed later in the day as dealers enter into repurchase agreements.

17

See New York Clearing House Association (1989), for the proposed loss-sharing formula for CHIPS.

18

A large proportion of the assets of a money center or a clearing bank are financed by short-term funds—certificates of deposit, repurchase agreements, interbank loans—and a loss of such funding might make it necessary for the bank to discount assets other than the eligible government securities. In this case the central bank would be exposed to the private credit risk inherent in such assets.

19

Foreign banks clear through a CHIPS settlement bank.

20

CHIPS operates under the legal environment of the U.S. Uniform Commercial Code and wire transfers therefore represent noncontingent commitments among banks.

22

An important category of transactions not yet subject to the daylight overdraft caps are book entry transfers of U.S. Government and federal agency securities. It is proposed that book-entry-related overdrafts also be included within the existing caps on depository institutions’ daylight overdrafts. Since the majority of book entry overdrafts appear to be concentrated in a few clearing banks that act on behalf of major dealers and brokers in government securities, it has been feared that a restriction on the overdrafts of such institutions could impair the smooth functioning of the Government’s securities markets. It is therefore proposed to permit such institutions to pledge book entry securities in transit as collateral for the amount of the total funds and book entry overdrafts.

24

Faulhaber, Phillips, and Santomero (1990) suggested that efficient pricing of overdrafts would involve a per unit price per transaction, a charge on daylight overdraft, and a premium covering possible default on daylight overdraft.

26

The development of multilateral clearinghouses could also significantly alter the structure of interbank credit relationships. For example, several large over-the-counter markets such as the interbank foreign exchange markets and the interbank swap market could evolve into an organized exchange, as has happened with Eurodollar futures markets. If this occurred, net claims on the clearing organization would replace gross interbank credit exposure in the deposit markets. Under the Basle capital adequacy standards, bank claims on organized financial exchanges subject to daily margining have a zero-risk weight in determining a bank’s required regulatory capital.

28

If a yen purchase occurs in New York and if the buyer would like to have the yen delivered during that day, for example, the Bank of Japan would have to make a transfer settlement at night. Under the current system, final dollar settlement in a yen-dollar transaction in the New York foreign exchange market is made through Fedwire during New York daytime, while the yen settlement is done through the GAITAME-YEN Settlement System during Japanese daytime, that is, early in the morning of the next day in New York. Hence, a time lag occurs between the two operations.

29

For example, Chase Tokyo operates a dollar settlement system in Tokyo. Transfers among the correspondent’s dollar deposit accounts with Chase are made in Tokyo. Balances are then transferred from Chase Tokyo to Chase New York after business hours in Tokyo. The final settlement occurs after the opening of the New York market. A net debit dollar position with Chase is financed by an overdraft extended by Chase Tokyo until the net debit bank is in a position to raise dollar funds in New York through CHIPS and finally deliver them through Fedwire. If the net debit bank in Tokyo exceeds its credit line with Chase, it will then have to raise dollar funds, which may be difficult when the Federal Reserve’s funds market is closed. If Fedwire operated on a 24-hour basis, it would become necessary to define when and how reserve positions are measured. Currently reserve balances are measured at the end of the day and counted as reserve balances for that day. In a 24-hour payment system, banks might be tempted to deposit funds for the measurement of reserve balances at a particular time and withdraw the funds immediately afterward. Another difficulty might occur in monitoring institutions. If, for example, a U.S. bank branch in Tokyo runs a negative balance on its accounts for the Federal Reserve, the latter may be unable to ascertain whether there is a problem, since the bank is located in Tokyo.