4 Large-Value Transfer Systems: Risk and Credit
- Omotunde Johnson, Jean-Marc Destresse, Nicholas Roberts, Mark Swinburne, Tonny Lybek, and Richard Abrams
- Published Date:
- March 1998
From the perspective of both monetary operations and prudential risk management, the priority for central banks’ payment policy has been the systems handling large-value payments. In this context, measures to reduce systemic risks in the payment system, and improve risk management generally, attempt to strengthen the incentives for banks to manage themselves prudently while reducing the public sector’s exposure to excessive risk.23 The risks typically emphasized in the context of payment systems are financial risks: liquidity risk, credit risk, and systemic risk. Liquidity risk is the probability that a payment will not be settled on time because the debtor has insufficient liquid funds. Credit risk is the probability that a payment will not be fully settled because the debtor is insolvent. Systemic risk is the probability that liquidity or solvency problems of one or more individuals or organizations in the payment system lead to liquidity or solvency problems on a large enough scale to threaten payment settlements in the economy at large. Apart from financial risks, there are also risks related to inadequate legal framework, human error, equipment failure, or low level of security of the system (so-called legal, operational, and security risks).
Systemic Risks, Settlement Finality, and the Central Bank
Systemic risks raise the greatest concern. Much of the discussion of those risks has focused on the effects of sudden and unexpected failure of one or more banks (participants) in LVTSs. For instance, simulations have been done, for both CHIPS in the United States and the Italian clearing system, of the effects on other participants in the system of a failure to settle by one participant, in the absence of a mechanism by the system to ensure settlement, and of intervention by the central bank. Although the simulations are not exactly the same in their methodology, they are reasonably comparable. They show that the systemic effects (other banks defaulting on their payments) could be quite sizable for CHIPS but rather small for Italy, both in terms of the relative number of banks affected and the relative monetary values involved. The differences seem to be due at least in part to the lower payments relative to GNP and higher degree of payments concentration in the Italian clearing system.24
Discussion of systemic risks has highlighted the central role of settlement finality in assessing those risks in a payment system. To ensure settlement finality, a system typically puts in place a mechanism whereby settlement can occur even in the case of failure of a participant. Settlement finality virtually eliminates systemic risk emanating from liquidity or solvency problems of participants; but it creates credit risk for the system and its settlement agent. Risk control policies therefore often involve measures to ensure settlement finality while simultaneously addressing the credit risk to the system and settlement agent associated with this assurance.25
Payment is final when it becomes irrevocable and unconditional. The central bank typically stands ready to provide final settlement facilities for private payment systems—both retail and large-value—subject to adequate safeguards to limit credit expansion. While banks can, as a matter of principle, settle using bilateral accounts with each other, or on the books of some private settlement agent (clearing bank), settlement by banks on the books of the central bank—whether gross or after multilateral netting—can be seen as facilitating a reduction of systemic risk. Payments using central bank money result in claims on the central bank which cannot fail (become insolvent) or have liquidity problems; from the perspective of agents other than the central bank, such payments, therefore, do not have any credit or liquidity risks associated with them.
Despite the special qualification of the central bank for handling payment settlement, it does not necessarily follow that the legal framework should require private clearing systems to settle their final obligations across the books of the central bank. Nevertheless, there is growing consensus for this approach toward clearinghouses and interbank net settlement systems, mainly because of the legal obligation of the central bank to promote monetary and financial stability.
Risk Management in Funds Transfer Clearing and Settlement
In funds transfer, there are basically four types of agents: the senders and receivers of funds (counterparties); the sending financial firms, typically banks (payment intermediaries); the clearing organizations (for example, clearinghouses); and the settlement agent (the clearing bank or central bank). Counterparties and payment intermediaries are often one and the same.
From the perspective of systemic risk, interest has primarily focused on risk management by clearing organizations and the settlement agent. Apart from strict membership criteria to ensure the soundness of the financial institutions involved, the measures that have emerged, especially since the early 1980s, to address risks borne by these organizations and agents can be divided into four types: exposure limits; collateralization; loss-sharing arrangements; and shortening of time lags in settlements.26
Exposure limits, commonly called debit caps and credit limits, are often used to contain systemic risks; these can be established as bilateral or multilateral limits. Among interbank large-value systems, these are fully developed (for instance, CHIPS in the United States and CHAPS in the United Kingdom).27 The net debit caps and net credit limits can, in principle, be absolute amounts or can be specified as multiples of capital. Ideally, for each participant there are at least two desirable controls: a net debit cap and a bilateral net credit limit. More specifically, there is, first of all, a network-specific net debit cap, defined as the maximum debit that a participant, by agreement, can have vis-à-vis other participants of the same system. Second, there is a bilateral net credit limit that specifies the maximum that a participant is willing to have as a net credit position vis-à-vis another participant. Each participant typically provides the system administrator with the bilateral credit limits it has established for all other participants.28
Collateralization and Loss-Sharing Arrangements
An important device used in interbank net settlement systems, to complete system settlement in the event that some participant cannot meet its net obligations, is to have participants post collateral as part of membership and to use this collateral to facilitate settlement. Typically, highly liquid assets such as government securities will be used as collateral, with such assets held in an account of the system at some securities depository (especially one managed by the central bank).
When collateral pledged by a defaulting participant is insufficient to ensure settlement within the system, ensuing losses to other participants could be distributed using various criteria in accordance with explicitly agreed loss-sharing rules. Generally such rules attempt to have losses distributed in relation to exposures to the defaulting participant. Both CHIPS and the Foreign Exchange Yen Clearing System (FEYCS) of Japan have explicit collateral-based loss-sharing rules.29
Apart from collateralization and loss-sharing arrangements, settlement failures can also be addressed through unwinding. An “unwind” occurs by deleting some or all of a failed participant’s obligations from a multilateral clearing and redoing the settlement calculations. Unwind solutions have become increasingly unpopular because of potential systemic disruptions. A preference for rules involving collateralization and explicit loss sharing has clearly emerged for interbank net settlement systems.
Time Lags in Settlement
Same-day settlement has also become a primary goal of clearing and settlement systems, since net obligations do not get carried from one day to the next, avoiding the risk that a participant with a large debit position could fail overnight or over holidays and weekends. All major interbank LVTSs have or plan same-day settlement.
Shortening the intraday transaction period could further reduce exposure time. In interbank net settlement systems, this could mean clearing and settlement several times during the day. This approach is not popular because the benefit in risk management may not compensate for the cost in additional reserve balances required relative to the alternative of daily netting with same-day settlement coupled with collateral-based loss-sharing arrangements. In RTGS systems, since settlement takes place transaction by transaction in real time, exposure time becomes virtually zero. Depending on how the system works—most notably, as regards intraday credit and the nature of any queuing—credit exposure gets shifted either to private sector agents outside of the particular organization or system or onto the central bank. When borne by the central bank, it can address the risks through use of quantitative limits on credit exposure, collateralization, and other devices.
LVTSs: Net Versus Gross
In a net or deferred-settlement system, payments sent and received among banks accumulate, and only the resulting net positions are settled at one or several designated settlement times during the day. Only one net obligation for each net debtor bank is due at settlement time. Netting reduces the size of the balances needed to settle large-value payments, but it also relies on an implicit extension of credit among participants between settlement times. The payments made between settlement times are not final and remain provisional until settled. In a gross settlement system, each payment is sent separately and is settled at the time it is sent, if sufficient funds for settlement are available. When associated with real-time processing, gross systems allow for real-time settlement and, therefore, for intraday finality.
Thus the choice between gross and net settlement in LVTSs depends on the relative costs and benefits in terms of settlement risks, operational efficiency, and reserve needs. The assessment of these factors, and hence the balance of considerations, is affected by whether the system is owned and administered by the central bank or by some private (commercial banking) organization.
The main advantage of a gross settlement system is that it ensures intraday finality. This both precludes accumulation of unsettled balances and the associated extension of interbank intraday credit (which occurs in netting systems) and helps to avoid the systemic disruptions that can result if a major segment of the system cannot settle at a prescribed time (say at the end of the day). Also, the central bank is relieved of the positive credit risk (however small) that it may be called upon to bail out a net settlement system that is in danger of causing systemic disruptions that can arise, for instance, when one or more of the participants in the system cannot settle, for liquidity or other reasons, and the system itself does not have adequate measures and resources to ensure settlement finality at the appointed time. Such an event could compromise the central bank’s monetary policy.
The main disadvantage of a gross system is the risk of gridlock from insufficient intraday liquidity (in terms of clearing or reserve balances) to ensure high operational efficiency of the system. This disadvantage has been addressed in various countries through some combination of queuing mechanisms, central bank intraday credit, and central bank reserve management policy.
The main advantage of netting systems is the saving in liquidity needed to support any given gross volume of payment transactions, which can be greatly beneficial to the operational efficiency of the system. The main disadvantage is the risk of settlement problems arising from settlement failure owing to liquidity or solvency difficulties of one or more participants. This disadvantage has been addressed by various methods to ensure settlement finality within netting systems without central bank intervention. To reduce systemic risks, net systems can impose membership criteria, shorten settlement lags, limit intraday exposures by imposing bilateral or multilateral limits, develop loss-sharing agreements in case of a settlement failure, or require the posting of collateral. Such procedures, along with other risk-management measures, have notably been developed by G-10 central banks under the aegis of the BIS. In particular, the Lamfalussy standards have emerged as the basic framework for assessing and regulating private netting systems.
It is, therefore, possible to opt for a gross settlement system, because of its advantages in the area of risk, and take appropriate measures to tackle the operational efficiency problems that can arise. This seems to be what is increasingly occurring when central banks own and operate LVTSs. Central banks in a number of countries are currently implementing RTGS systems: among others, the European Union central banks have decided that each country should have an RTGS system for large-value payments and that domestic RTGS systems will be linked to build a Pan-European RTGS system. By the same token, it is equally possible to opt for a netting system, in light of its economical requirements for reserve balances, and take appropriate measures to ensure settlement finality within the system. This seems to be the tendency when the LVTS is owned and operated by the private sector. Thus, it is not an exaggeration to conclude that a major reason for the tendency toward RTGS systems is that, apart from some notable exceptions (CHIPS in the United States, CHAPS in the United Kingdom, UPS in Canada, Bill and Cheque Clearing System in Japan, and BGC-SWIFT in the Netherlands), LVTSs tend to be owned and administered mainly by central banks. If central banks want to encourage the private sector to embrace gross systems, because of a strong view that gross systems are superior to netting systems from the point of view of global welfare, then the central banks must find a way—hopefully short of direct regulation or inefficient subsidization—to assure the private sector that adequate liquidity (that is, base money) will be supplied by the central banks to the economy as a whole to guarantee a level of operational efficiency of the system no less than what would be attained under a netting system.
In brief, the trade-offs between real-time gross and net settlement systems may be viewed differently by the commercial banks and the central banks. From a commercial bank’s perspective, one of the key questions is the opportunity cost of reserve balances held for settlement purposes. From the central bank’s perspective, the stability of the payment system is a critical issue, and most central banks tend to perceive RTGS systems as associated with a level of systemic risk that is lower than that of net settlement systems. However, interbank daylight credit (and its associated systemic risks) can also effectively emerge in RTGS systems, depending on the message flow structure (discussed below).
Central Bank Credit Policy for RTGS Systems
The current global trend toward RTGS systems has heightened the relevance of the question of whether the central bank should grant credit directly to support such systems.30 When the central bank provides such credit facilities (typically intraday credit facilities), a number of questions arise, including: (1) whether and why any intraday credit should be granted by the central bank; (2) the kind of intraday credit that should be granted; and (3) what should be the indicators for determining interest and other charges on the credit.
Whether Intraday Credit Should Be Granted
The value of funds transfers that occur during any single day is typically several times the underlying bank reserves available for final settlement. Averaging of reserve holdings for purposes of meeting required reserves and permitting intraday use of all reserve balances for payment settlement purposes can help in easing the additional pressure for intraday reserves associated with a move from end-of-day net settlement to RTGS. But this may still not suffice to ensure tolerable operational efficiency of the system. Thus, bank reserves must turn over several times during the day (hence the notion of “turnover ratio”) if settlement is real-time gross.31
For any given level of reserves, it may be possible to increase the turnover ratio and thus support greater gross transfers. For instance, certain payments could be delayed (“delayed sends”) until covering funds become available. This sort of queuing of payments could be centralized—by having the payments organization or system doing the queuing, as in the SIC system (see “Payment Systems in Switzerland” in BIS (1993b), pp. 351-83; see also the Appendix, Table A19)—or it could be decentralized by having the sending financial firms do the queuing.32
A major issue for the central bank is that it will not want its credit operations in the payment system to reduce its control over liquidity (and monetary) management. Also, it will want to avoid the moral hazard of the payment system users viewing the central bank as a lender of first resort rather than using private money markets as much as possible.33 These and other considerations imply that the case for direct central bank intraday credit to support an RTGS system is strongest when such operations are not expected to conflict with the central bank’s basic objective of controlling inflation (perhaps because the bank is able to institute effective mechanisms to control such credit), and when the private money markets are not in a position to adequately satisfy the credit needs for a smooth operation (operational efficiency) of the system without periodic stress. Of course, in practice, there are also various historical reasons why countries may or may not be providing intraday credit for RTGS systems. Among central banks that currently provide intraday credit for their RTGS systems are Denmark, the United States, and Thailand, while Germany, Japan, and Switzerland are among those that do not (see the Appendix tables).
Types of Intraday Credit
Daylight overdrafts can be provided via collateralized or uncollateralized credit, sometimes up to some maximum amount (especially if uncollateralized); the collateralized credit can take the form of an intraday repo facility. Whether or not the central bank has intraday credit facilities, the bank can also have its other short-term credit facilities (overnight loans, Lombard facility, discount window) used to assist banks in their RTGS payment settlements near the end of the day; this could be especially valuable in queuing systems.
It has been proposed that central banks either charge an explicit interest or fee for use of intraday credit34 or else require supplemental reserve balances linked to the volume of the overdrafts (see Humphrey, 1992; Hamdani and Wenninger, 1988; and Belton, Gelfand, Humphrey, and Marquardt, 1987). In this vein, the United States Fedwire now charges an explicit fee for daylight overdrafts.35 On April 4, 1994, an effective daily fee of 10 basis points (annual rate) went into effect; the fee was raised, effective April 13, 1995, to 15 basis points. According to the U.S. Federal Reserve the daylight overdraft fees have had a substantial dampening effect on overdraft levels. Daily peak overdrafts, which averaged $185 billion in 1994 before daylight overdraft fees went into effect, fell to an average of $145 billion for the rest of the year; the average per-minute overdrafts dropped by 37 percent. The reduction in overdraft was not accompanied by any perceptible market disruptions.36 The different ways in which users of daylight overdrafts could have adjusted to the U.S. Federal Reserve pricing policy to limit daylight overdrafts have been analyzed in the literature, but the relative contribution of these alternative organizational practices are not known.37
Pricing of Intraday Credit
An issue of practical importance involves finding indicators to use as benchmarks for the pricing (supply price) of intraday credit. In principle, this price would be equivalent to the sum of the administrative costs, the opportunity cost of the funds, and some adjustment for risk. If a market existed and all risks were internalized, this risk adjustment would include the credit risk connected with the overdraft. In the absence of an actual intraday private market for funds, such as the intraday market in Japan, a central bank must look for proxies. A reasonable approach would start with the interest rate prevailing in a fairly similar funds market and adjust it for risk, maturity, and implicit charges for other services jointly supplied.38 Another approach would start with some central bank overnight interest rate, adjusted for maturity and for the extent of relative collateralization. In Thailand, for example, the interest rates for borrowings under the Intraday Liquidity Facility of the Bank of Thailand are linked to the previous day’s repurchase market rates.
Interday and Intraday Liquidity
It was stated earlier that aspects of the decision-making process in monetary policy need to be reexamined when reforms or endogenous changes are taking place in the payment system, namely: the monetary policy target and instrument settings, the choice of appropriate target or indicator variables, the design of monetary policy instruments, and the monetary policy transmission mechanism itself. In this context, there is some more formal analysis to suggest that, in principle at least, the intraday and interday markets cannot be neatly segmented and that movements in intraday conditions and rates will indeed influence interday rates (see VanHoose, 1991). This would imply that, particularly as intraday markets develop, central banks may need to take this into account as they formulate their monetary policy, including the role of various interest rates as policy targets and instruments. In practice, so far, central banks generally seem to have taken the view that, while the links between interday and intraday liquidity may exist in principle, quantitatively they are not yet of major importance.39 For the time being at least, monetary policy under existing RTGS systems still effectively operates in practice on end-of-day balances and interday/overnight interest rates.
Potential Effect of Intraday Credit on Financial Markets
When RTGS systems are introduced and intraday liquidity is limited, it may affect the trading patterns in certain financial markets during the day and, thus, liquidity; this may be reflected, for example, in the bid-offer spread in the financial markets. The recent experience of the United States provides a good illustration. In the United States, where a significant share of the intraday overdrafts are related to trading in government securities, the introduction of a fee for overdrafts apparently significantly changed the trading pattern during the day. U.S. government securities dealers began arranging their financing transactions earlier in the morning and delivering securities used as collateral for repos more quickly to their counterparties to cover overdrafts caused by early morning repayment of maturing repos.40 Also, to the extent that intraday credit has not previously been collateralized and collateral is then required, as in some countries in the European Union, the demand for securities eligible as collateral will, all other things equal, tend to increase (see Folkerts-Landau, Garber, and Schoenmaker, 1996).
Regulation of Private LVTSs
The central bank will typically have regulatory powers with respect to private participants in the payment system. As regards risk, one type of such regulation which a central bank could enforce has been mentioned above, namely, requiring that certain clearing organizations and interbank large-value (net) settlement systems settle in the books of the central bank. A second type of regulation addresses risk management directly (in a manner described earlier in the chapter), while a third type of regulation sets out licensing and reporting requirements for engaging in certain activities in the payment system. In the area of regulation, central banks increasingly seek, inter alia, to ensure that the six Lamfalussy standards are met by interbank net settlement systems.
With regard to credit, the question is whether the central bank should control lending by organizations in private payment system networks—especially daylight credit. If such lending is explicit, in the sense that it involves a transfer of reserve balances (at the central bank) at the time of the lending, there may be no strong case to control it in the context of payment system regulation. The issue is more important when the lending takes the form of debits by the debtor organization (mainly banks) and “due from” is built up by the lending organization. In general, the case for central bank regulation of such debit/credit positions would be strongest when the central bank is likely to feel compelled to bail out the organizations in the case of a systemic crisis.41
Varieties of LVTSs
As noted earlier in this chapter, there has been a growing focus on large-value payments. The Appendix toward the end of the book provides examples of systems around the world handling the clearing and settlement of such payments. The systems reflect the heterogeneity of the countries chosen, with regard, inter alia, to the level of development, history, banking and financial structure, and the role of the central bank in the payment system. However, all the systems included in the sample have a fair degree of specialization in handling large-value payments. This section summarizes critical LVTS features—some of which have been mentioned earlier—that would facilitate understanding of the tables in the Appendix.
The key features of LVTSs can be classified under: (1) general organization; (2) clearing and settlement cycle; and (3) risk control measures. Risk control measures have already been discussed. In addition to discussing the other two features, this section also introduces the important issue of RTGS message flow designs.
Organization deals, first of all, with ownership, operation and management, and membership of the systems. For instance, a system may be owned by the central bank, a commercial bank, groups of commercial banks, or a payment association; it could be governed by a committee, the chairman of a payment association, or representatives of the central bank. LVTSs can also be organizationally described in terms of membership criteria, types (banks or nonbanks), and number of current participants, as well as in terms of any tiering arrangements.
In LVTS design, various other basic features must be decided in the organization. One is account structure. This has to do mainly with the number of accounts per participant, with the settlement agent, and the number of processing sites (for payments clearing). Another basic feature is whether there will be limitations on the value of transactions to be handled by the system or whether participants will be free to make transfers of any size as long as they are willing and able to pay the transaction fee. A third issue is the nature of the settlement (for example, net or gross or both), and a fourth is the fee structure and the underlying principle governing this structure (for example, full cost recovery). Finally, there is the important question of the legal/contractual framework governing the system. For instance, the legal basis could be a law on the payment system, central bank regulation, central bank law, or a contractual agreement (enforceable in the courts) between the users themselves and between the users and the central bank.
Current LVTSs have evolved in response to requirements for speed, security, operational efficiency, reliability, and risk management. Specific choices in design features have been shaped by the above general requirements—and hence the circumstances of the countries. But specific design choices have also been shaped by the experience and history of the country as regards payment systems, financial markets (including systemic crises), and monetary policy, as well as by the budgetary constraints faced by the country. For example, some systems are totally new, while others have evolved from preexisting systems, generally by creating or improving their risk management features and upgrading their technical capabilities (in terms of processing speed, operational capacity, real-time monitoring of accounts, and delivery of information to the participants).
Clearing and Settlement Cycle
Aspects of the clearing and settlement cycle of an LVTS have to do with issues such as the type of instruments handled by the system; the timetable for the different operations in the cycle; the nature of any queuing mechanisms in place; liquidity and credit facilities (discussed at length before); the information (or message) flow in the system; time of settlement finality; and whether or not the particular LVTS system will be connected with other domestic systems. The information (or message) flow designs are discussed later.
The types of instruments to be handled by an LVTS could be paper-based, electronic, debit, or credit.
RTGS will have implicitly or explicitly certain queuing features. Some systems are designed with centralized (within the “system”) queuing mechanisms for settlement, while in other cases payment requests may simply be rejected and returned to the sending bank when no funds are available for settlement; in the latter cases, the participants (banks typically) have to do their own queuing (namely, a decentralized queuing mechanism is in operation). When queuing is centralized, payments may be settled on a first-in, first-out basis, which may or may not be modified by a system of priorities; for example, certain types of payments could be considered top-priority payments. Such priority could even be determined by official regulation.
As regards time of settlement, as discussed earlier, this could, for example, be real-time, at designated times during the day, end-of-day, or next-day. Finally, an LVTS will normally be connected operationally with retail payment systems or with domestic securities or foreign payment systems.
RTGS Message Flow Design
Figure 1.RTGS Message Flow: “V” Design Figure 2.RTGS Message Flow: “Y” Design Figure 3.RTGS Message Flow: “L” Design Figure 4.RTGS Message Flow: “T” Design
Figure 1 shows the “V” design. The sending bank sends the payment instruction to the central bank, which in turn sends it to the receiving bank only after payment settlement (after the sending bank’s account has been debited and the receiving bank’s account has been credited). The central bank is at the center of the information flows, and it receives and sends the full payment messages.
In the “Y” design (Figure 2), a central processor, located at the joint of the “Y,” strips the payment instruction received from the sending bank by suppressing all the commercial information not strictly needed for settlement. It then sends a settlement request to the central bank, and the payment instruction is retained by the central processor until confirmation of settlement is received from the central bank, at which time the full information is delivered to the receiving bank. The central processor is at the center of the information flows, but the central bank, by sending the confirmation, initiates the sending of the payment instruction by the central processor to the receiving bank. In terms of flows of information, the central bank receives only a settlement request (not the full payment instruction) and sends a settlement confirmation.
Figure 3 shows the “L” design where the payment instruction is retained by the sending bank’s local processor (also called gateway) until confirmation of settlement is received from the central bank. After confirmation, the payment instruction is released to the receiving bank. The central bank initiates the sending of the payment instruction by the sending bank’s gateway to the receiving bank. The central bank receives only a settlement request (not the full payment instruction) from the sending bank and, once it is processed, sends in return a settlement confirmation.
As shown in Figure 4, under the “T” design, the sending bank sends, simultaneously, the payment instruction to the receiving bank and to the central bank. Therefore, the receiving bank usually receives the payment instruction before receiving confirmation of settlement by the central bank.
Message Flow Design and Risk
The different message flow designs have implications for risk. The “T” design allows a payment instruction to be automatically sent to the receiving bank before it is final, that is, before it is accepted and settled by the central bank. This information is very valuable for the banks because they can use it for intraday cash management purposes. However, if a receiving bank acts on this information, in its liquidity management, on the assumption that it will receive the funds before the end of the cycle, it would become exposed to liquidity risk. If, in addition, the receiving bank gives third-party customers use of the funds before settlement—for instance, by allowing a cash withdrawal—it will run a credit risk vis-à-vis the sending bank. Allowing customers advance use of funds may result, for example, from commercial pressures (competition) or from contractual or regulatory provisions obligating a bank having received a payment through a given system to credit the customer’s account immediately. Both credit and liquidity risks can be significant sources of systemic risk. For these reasons, “V,” “L,” and “Y” designs are considered safer than the “T” design. Not surprisingly, therefore, most of the RTGS systems in operation or under development use one of the first three designs, and the “T” design, which was initially a standard product for RTGS systems developed by SWIFT, has been largely replaced by the “Y” design.
However, it is worth noting that, irrespective of the message flow design, banks can have access to information on unsettled incoming payments through the mechanism of pending queues. These mechanisms allow temporary unsettled payments to accumulate in a certain order during the day, over a certain period, until sufficient funds are available in the sending bank’s accounts. If the receiving banks have real-time access to the awaiting incoming payments which have been sent to them but not yet settled—that is, that the pending queues are “transparent,” their behavior may result in risks very similar to those arising under a “T” system.
The access to information on queued incoming payments is a controversial issue in the discussion on RTGS designs, mainly because of the effective trade-off between financial risk and operational efficiency. The access facilitates better liquidity management by banks, but it also provides an incentive for banks to take additional risks. However, even while sometimes limiting the content of the information or making it available only on request, most of the RTGS systems with queuing mechanisms allow receiving banks access to the queued incoming payments.
The great variety of existing LVTSs is evidence both that no single model or approach is necessarily the best under all circumstances and that the dynamic path of evolution of a system helps to explain the current design of that particular system. The focus on LVTSs is in many countries relatively new and many of the systems described in the tables are not operating yet. However, it is possible, despite the diversity of the systems under review, to highlight a few common, or at least dominant, trends.
The tables of the Appendix show that most countries have two or more systems handling large-value interbank payments. In some countries, this situation is simply due to the fact that the specialization process is just beginning and that, since no specialized system exists, large-value payments are handled within the existing interbank transfer systems. But in a few other countries (for example, the United States or Spain), large-value payments are—or will be—handled by at least two specialized systems: a multilateral net settlement system, generally operated by commercial banks, and an RTGS system managed by the central bank. Available data show that the total amounts of payments exchanged through specialized large-value systems are usually very large: in many systems, the daily flows equal the value of annual GDP every few days.
As for the type of LVTS, all but four countries among those in the Appendix (Canada, Israel, India, and Malaysia) already have, or are developing, an RTGS. For the European Union countries under review (Denmark, France, Germany, Italy, Netherlands, Portugal, Spain, Sweden, and the United Kingdom), this development is part of the ongoing harmonization process between their domestic payment systems.
The central bank is the settlement agent for all the LVTSs under review. When commercial banks operate systems (usually net) that coexist with central bank-operated RTGS systems, the private systems usually settle through the RTGS. This common feature illustrates clearly the attractiveness and/or necessity of central bank money in settling interbank transactions, given the finality it confers on settlements.
Half of the countries reviewed impose a minimum amount on the transactions processed by the LVTS. However, this restriction very often applies only to transactions initiated by customers and not to pure interbank transactions. The LVTSs process a wide range of payment instruments, and even in some cases checks. However, as the systems tend to specialize, and especially when RTGS is operating, electronic credit transfers are by far the dominant instrument. All the systems function on a full or partial cost recovery basis; in other words, no system is completely free of charge to its users.
There is quite a variety in the legal and contractual framework governing large-value transfer systems. However, in most cases, no specific law has been enacted, and the systems operate within the framework of contractual agreements concluded between the commercial banks and the central bank.
The provision of intraday liquidity is especially relevant in the design of an RTGS system. Among the RTGS systems reviewed, (1) five central banks (in China, Germany, Japan, Korea, Switzerland) do not provide any intraday liquidity; (2) ten central banks provide it either through collateralized overdrafts (Denmark, Italy, Netherlands, Portugal, Spain, Sweden, Thailand) or through intraday repos (Australia, France, the United Kingdom); and (3) two central banks provide intraday uncollateralized overdrafts (the United States and Mexico). However, when no intraday liquidity is provided by the central bank, all systems, except in Japan, have sophisticated queuing mechanisms.
In the case of net systems, where management of settlement risks is especially of serious concern, five systems (in Canada, India, Mexico, Spain, and the United States) rely essentially on a combination of limits and loss-sharing agreements; Japan and Malaysia have no specific risk control measures. As for the message flow design, most of the RTGS systems reviewed use a “Y” or “V” design. Only one “L” and one “T” design are found—in the United Kingdom and Sweden, respectively.