Printer Friendly

International settlements: a new source of systemic risk?

The very real significant social costs of systemic risk have long served as

an important rationale for a federal presence in the domestic payments

system.(1) Recent market developments have heightened concerns about the

potential for systemic risk in the payments system. First, the sheer growth

in large-volume payments has raised the potential, costs should a number

of institutions fail. Second, technology and technological change seem to be

redefining the kinds of transactions taking place as well as increasing the

speed with which these transactions can be completed and funds transferred.

For example, both computer and options pricing technologies now permit the

unbunding, restructuring, and creation of transactions (such as swaps and

derivatives) whose risks, legal status, and related characteristics are just

now beginning to be understood.

A third dynamic behind the increased concern about

systemic risk in the payments system is the globalization of

financial markets, which is tying economies and

markets together in ways that introduce additional

issues about the mechanisms by which traditional

clearing (the notification and transfer of documents and

orders to purchase and sell assets) and settlement (the

transfer of final payment) take place. Finally, the fears

about the supposed potential for systemic risk associated with

clearing and settlement loss have been given greater

credence by the lack of internal controls within major

institutions, which have been exposed by the actions of

rogue traders in Kidder, Bankers Trust, and Barings

(see Edwards 1996). As the problems in these institutions

have been unwound, greater appreciation has emerged of just

how complex and segmented the institutional arrangements

for clearing and settling transactions have become.

Despite the fact that securities, futures and options, and

derivatives are increasingly cleared under a variety of

institutional arrangements, final settlement usually takes

place in the interbank market. In general, clearing of

transactions--be they securities, derivatives, or other assets--is

almost exclusively done by the private sector while

settlement can take place in the wholesale banking sector or

through central banks (see BIS 1997b).(2) Markets have

become tiered as more and more transactions are cleared

through several layers of institutions before they are

ultimately settled (see Corrigan 1990). Equally important,

the introduction of new instruments, such as swaps,

collateralized mortgage obligations, and off-exchange

derivatives, and their associated methods for transferring

cash flows and settlement relationships have resulted in

seemingly unrelated markets and institutions being linked

together in ways that both create and may de facto transfer

risks from one market to another. Increasingly, these

transactions and markets are assuming an international

dimension that can also have significant domestic market

implications (see BIS 1997b).

This article examines whether internationalization has

changed the nature of and potential vulnerability of the

financial system to systemic risks and looks at a method to

mitigate them. The Lamfalussy Report, which examined and

proposed standards for payments systems settlement and

risk control features, indicates that system vulnerability is

critically linked to the length of time that participants are

exposed to credit and liquidity risks (see BIS 1993b). The

analysis presented here suggests that regulatory and legal

structures can also have systemic risk dimensions. As to the

fundamental question of whether new risks are being

introduced, the answer seems to be no. Moreover, recent

institutional and regulatory developments may act to reduce

the potential scope and the size of these risks and limit

implicit taxpayer liabilities should these risks be realized.

Risks in Payments

Regardless of the institutional arrangements, there are four

generally accepted generic types of payments system risks

that have been identified and have been the focus of much

attention. These include operational, legal, credit, and

liquidity risks (see Eisenbeis 1995 or BIS 1997b). While it

is easy to differentiate these risks conceptually, in reality

they tend to be interrelated. The realization of one can lead

to occurrences of the others, and this dynamic has not

changed with the evolution of the new instruments and

markets just described. These interrelationships among risks can

be illustrated by considering credit risk, which arises when

the purchaser of an asset defaults by failing to settle any or

all of its obligations. Credit risk arises as a logical by-product

of separating the clearing and settlement functions, which

under current institutional arrangements nearly always

involves an extension of temporary credit.

Credit risk is a function of the potential loss exposure

when a buyer initiates a transaction ordering its bank to

transfer funds but then cannot make payment without going

into an overdraft situation. The buyer's bank, which is

attempting to settle on behalf of the buyer, is faced with

essentially three alternatives. First, it can provide credit to

the customer until funds are received. Second, the transaction

can be canceled, or the bank can complete the transaction

itself. If the buyer's bank takes the place of the customer and

completes the transaction, it may then take possession of

the goods or asset (or any other of the customer's available

collateral) and proceed to unwind the transaction. Finally, in

the extreme, the buyer's bank can default on its own

obligation to settle if the time for settlement has not yet


If the buyer has good collateral and a sound credit

rating, then extension of credit may be the best alternative.

Canceling the transaction may not be an option, especially

when delivery of the good or service has already taken place

and there is no available collateral.

Settlement failure in this example could be controlled if

the buyer's bank were to put a hold on the buyer's funds at

the time payment is initiated, collateralizing the transaction.

Organized futures markets effectively accomplish this

control through the use of margins, mark-to-market

accounting, and settlement requirements. For good

customers, however, collateralization may not be necessary,

practical, or efficient, especially if both the probability of

default and the expected loss are small relative to the bank's

resources. The lack of a hold or similar type of collateral

policy illustrates that an institution's vulnerability and

exposure to credit risk often results from the underlying

conventions, practices, and

structures of the markets involved rather than from the

realization of performance risks associated with the

underlying projects and investments.

As markets have become increasingly global, differences

in timing and clearing and settlement conventions and

differences in bankruptcy laws can add important temporal

and other dimensions to credit risks not always found in

domestic markets. This consideration was clearly

demonstrated in 1974 when Herstatt Bank failed and was

closed by German authorities. Herstatt had entered into

agreements to exchange deutsche marks for dollars. The mark

leg of the transaction was settled, but the dollar portion was

not settled in New York at the time Herstatt was closed

since the deadline on CHIPS (Clearinghouse Interbank

Payments System) for final settlement was approximately

4:30 p.m. eastern standard time. This difference in settlement

times for the two sides of the transaction left the

counterparties to the foreign exchange transaction thinking

that they had more funds than they did. When the dollar

transactions failed to settle, the result was large losses to the

U.S. counterparties. This temporal dimension to

credit/systemic risk has come to be known as Herstatt risk

and can be very large.(3)

A more recent example of this type of event is the

closing of the Bank of Credit and Commerce International

(BCCI) in 1991. The Industrial Bank of Japan had paid 44

billion yen into BCCI's branch in Tokyo, for which payment

was to be received in New York from BCCI's New York

branch. When BCCI was closed, the dollar portion of the

transaction was never completed, and Industrial Bank of

Japan became a creditor for $30 million.

These examples may at first look like ordinary credit

risk in that loss exposure resulted from the inability of

Herstatt and BCCI to pay. But the incidence of the losses

and ultimate position of the banks' creditors was determined

by both home country laws and the intervention policies of

their regulatory authorities, whose actions usually cannot be

easily predicted or priced.(4) The losses to dollar

counterparties in the Herstatt case were the consequence of

the timing of the closure of the institution rather than the

realization of estimable default risk. Had the German

authorities waited until the U.S. dollar markets had settled,

then the losses to those expecting

dollar transfers would not have occurred and the risks would

not have been realized. Such exposure is better characterized

as settlement uncertainty rather than settlement or credit risk

since it is not possible to estimate reliably and cost out the

implications associated with the vagaries of sometimes

untested statutes governing transactions and of regulatory

actions and policies. Note, too, that although the size of the

losses may not have been affected by the closure timing, the

distribution of the losses was significantly affected by legal

structures and governmental action. At the same time,

numerous initiatives by governmental bodies such as the

Federal Reserve and the Bank for International Settlements

BIS are continually seeking to identify and institute

policies to limit these problems (see Bank of England 1994

and BIS 1989, 1990, 1993a-c, 1997a, b).

Herstatt-type risk can also be involved solely in dollar

clearing systems. In Asia the Chase Manhattan Bank

operates a dollar clearing and settlement service through its

Tokyo branch. The system provides a limited overdraft

facility and promises finality of settlement guaranteed by

Chase Manhattan. Participants are permitted to settle

overdrafts in New York across the Tokyo/New York

business day. Furthermore, Tokyo balances at the end of the

day may be transferred to New York through the New York

offices of Chase or Tokyo banks or through CHIPS. In this

system any problems that may arise in this satellite

settlement and clearing system quickly have the potential to

transmit liquidity and credit risk from Asia to New York,

and ultimately to the Federal Reserve, if it affects CHIPS,

Chase, or significant New York correspondents. A failure to

settle in New York on payments guaranteed in Japan by

Chase creates a form of Herstatt risk that would end up

having to be resolved in New York. At present, concern

about such clearing and settlement systems stems from the

sheer size of the potential losses rather than from a true

understanding of well-articulated scenarios on how the risks

would be played out.(5)

Sources of Payments Uncertainty

Whenever clearing and settlement of financial assets are

separated in the international arena, a given country's rules

usually establish the exact point in time that a transaction

has been completed and the obligation satisfied. The issue

centers on transaction finality and the legal criteria for when

debts are discharged and who bears the losses in the event of

default. Finality usually occurs when the party selling the

asset actually has "good funds" and the transaction is both

irrevocable and unconditional. Importantly, since many

central bank settlement systems can involve the extension of

intraday credit, finality may or may not correspond to the

time that the buyer actually settled. For example, because

Fedwire provides finality as a matter of Federal Reserve

policy, acceptance of a payment order

carries with it the "guarantee" of good funds to the receiver

and also discharges the debt, since the sender's reserve

account is debited and the receiver's bank account is credited,

even though the sender's bank may default on the settlement

of its reserve account with the Fed at the end of the day.

When the settling institutions are located in two separate

countries, the specifics of the transactions in terms of

settlement, discharge of debt, and so forth may sometimes be

governed by the laws of two separate countries and, if

transactions involve clearinghouses, the laws where they are

located as well.

The legal status of claims can quickly become very

murky when the problems involved in settlement failures in

cross-border bilateral and multilateral netting arrangements

are examined, especially those transactions involving

forward-dated contracts in foreign exchange, derivatives, and other

cross-border markets (see BIS 1997b). Under netting

systems, debt and credit orders are cumulated, and only the

net difference is transferred at an agreed-upon time. This

procedure contrasts with real-time gross settlement systems

(RIGS) which continually process and settle transactions as

the orders are received. Final disposition of the liability

under netting systems depends critically on the legal rules

governing the disposition of debts and transactions in the

event of a default or bankruptcy.

As an example, if two institutions have entered into a

bilateral netting arrangement, then completion of all the

transactions subject to the arrangement is contingent on

settlement of the net position. Should one of the parties fail

to settle because of a bankruptcy, all the gross transactions

subject to netting may have to be undone. The determining

factor here depends upon the legal rules affecting the markets

in which the transaction was settled. Since the legal rules

may differ according to where settlement takes place, and

this location may be beyond the receiver's control,

settlement uncertainty may exist.

The exact status of cross-border transactions, therefore,

is determined by several sets of laws. These include

the laws governing bilateral netting arrangements and those

governing the particular settlement market involved as well

as the bankruptcy provisions and other related laws of the

country of the failed institution (or the laws of the resident

country if the transaction is recorded on the books of a

branch of the failed bank). For example, netted transactions

may or may not be regarded as discharged. The bankruptcy

court with jurisdiction over the transactions may decide to

unbundle netted transactions, demanding payment for debts

owed and disavowing liabilities to creditors. In addition,

country bankruptcy law may give creditors the right to

offset their liabilities to a failed entity against their claims on

that entity. Thus, debts owed on foreign exchange may be

discharged with debts on securities, loans, or any other

assets. Not only do the bankruptcy laws affect the size of

the losses but also the way in which the losses may be

apportioned across various creditors.

The legal situation in multilateral netting

arrangements introduces complexities several orders of

magnitude greater than those affecting bilateral arrangements.

There is considerable variation across countries in treatment

of transactions, and thus uncertainty exists about how

particular bankruptcies will be treated. The key point is that

this legal uncertainty often can undermine the efficiency of

bilateral and multilateral netting arrangements and creates the

very real possibility that systemic risks could be heightened

rather than reduced when the laws governing netting are not

uniform across countries. Because these legal uncertainties complicate

assessment of the likely outcome of a default scenario for

many transactions, authorities have paid great attention to

putting transactions on a common legal basis and, as

discussed in the next section, some nations have moved to

establish real-time gross settlement as the basis for clearing

and settlement.

Responses to Uncertainty

Both private- and public-sector entities have responded to

the increased uncertainties, market risks, and evolving market

technologies in many interesting ways. The responses affect

contract design and the micromarket structure of exchanges

and their rules governing transactions. They have given rise

to proposals to change laws governing transactions and

suggestions to increase governmental cross-border

cooperation in financial rules, regulation, and supervision

as well as changes in the structural design of transfer


Given the complexity of financial transactions and their

interrelationships, measuring, monitoring, and pricing what

institutions' true risk exposures to each other are and how

these risks flow directly and indirectly through

relationships with related customer groups is difficult. For

example, Customer X may have several relationships with

its primary bank (Bank A). These might include a loan, a

swap, a deposit account, and several foreign exchange

transactions. Customer X may also have similar

relationships and transactions outstanding with Bank B. In

addition, Bank A may also have made loans in the form of

advancing federal funds to Bank B. If Customer X fails, the

entirety of its net position with Bank A across all the

relationships and transactions represents its net direct risk

exposure. Bank A may also be indirectly exposed through

Bank B if the customer's default causes Bank B to default on

its federal funds obligations to As primary bank.

Measuring and monitoring these interrelated exposures

across the world, across different markets and time zones, is

a truly daunting modeling and monitoring problem. It is made

even more so by the dynamic and continual evolution of new

instruments and markets.

Central bank and market responses to these challenges

have been to substitute rules and other mechanisms to

control customer risk-taking incentives. A

number of control mechanisms have been designed to limit

uncertainty and to provide incentives for member

institutions to control their own risk exposures. These

include maintenance of adequate capitalization, reliance upon

contract design to allocate risk and losses, collateralization of

transactions, use of outside guarantees and bonding, pricing,

imposition of system membership requirements, and

self-imposed (and system-mandated) caps and other limits on

risk exposure to individual and related parties. For example,

in the United States, the Federal Reserve imposed limits in

1986 on participating banks' net exposures across Fedwire

and CHIPS as well as bilateral limits on exposures to

individual participants. Collateralization of certain positions

is also required, and the system charges for intraday credit

that is extended.

Contracting activities also have focused on apportioning

risks, defining performance, and allocating losses among

participants in a payments system or exchange in the event

that a default occurs. Because of the difficulties in

continuously measuring and monitoring total risk exposure

to individual system members, caps on the amount of

exposure with any member have been imposed, and the

system imposes a similar total cap across all system

members. In the case of the U.S. CHIPS system (which is

not a real-time gross settlement system), participants require

same-day settlement, engage in real-time monitoring, have

established limits on exposures, have required collateral to

cover the largest two exposures, and have instituted a

loss-sharing arrangement.(6) System members also impose

various types of membership and participation

requirements, such as the maintenance of minimum capital


It has also been recognized that accounting rules--such as

mark-to-market requirements--can affect the ease of

information transfer and reduce monitoring costs. Such rules

have been especially widely used in the case of futures,

options, and commodities exchanges.

Finally, systems are evolving toward real-time gross

settlement despite the supposed efficiency advantages of

netting arrangements. Real-time gross settlement systems

require those engaging in payments activities to collateralize

payments fully as they are initiated. The benefits of doing so

are weighed against the costs of uncertainty and credit risks.

Such systems contain inherent incentives for institutions

engaged in offering payments services to price and monitor

their exposures. Furthermore, real-time gross settlement

reduces risk exposure by limiting the duration of both credit

and liquidity risk.

The first real-time gross settlement system was the

Federal Reserve's Fedwire (see BIS 1997b). By the end of

the 1980s, six of the Group of 10 countries had instituted

(RIGS) systems. As the European Union proceeds,

Lamfalussy Standards BIS 1993b) specify that (RIGS)

systems must be in place, and the union's umbrella

settlement system, Target, which will link settlement

systems within the union, is also designed as a real-time

gross settlement system. The progress of the European

Union and the European Monetary Union have also

contributed to the conversion of netting systems such as

the U.K. CHAPS system to real-time gross settlement even

though the United Kingdom is not projected to join the

European Monetary Union initially. Table 1 briefly

summarizes some of the salient characteristics of settlement

systems in selected developed countries and illustrates the

extent to which they are evolving toward real-time gross



Features of Selected Funds Transfer Systems

 System Central Bank

Country (Planned) Type Date Daylight Credit

Belgium ELLIPS RIGS 1996 Yes

Canada IIPS Net 1976

 (LVTS) Net 1997

France SAGITTAIRE Net 1984

 (TBF) RIGS 1997 Yes

Germany EIL-ZV RIGS 1987 Yes

 EAF2 Net 1996

Italy BISS RIGS 1989

 (BI-REAL) RIGS 1997 Yes

 ME Net 1989

 SIPS Net 1989

Japan BOJ-NET Net+RTGS 1988 No

 FEYCS Net 1989

Netherlands FA RTGS+Net 1985

 (TOP) (RIGS) 1997 Yes

Sweden RIX RIGS 1986 Yes

Switzerland SIC RIGS 1987 No

United Kingdom CHAPS RIGS 1984 Yes

United States CHIPS Net 1970 No

 Fedwire RIGS 1918 Yes

Source: BIS (1997b).

Conclusions and Implications

The present path on which payments systems are moving

involves a seeming contradiction. On the one hand, markets

are becoming more integrated

and global in scope. At the same time they are becoming

more segmented in the sense that there is a growing

separation evolving between the clearing and settlement of

transactions. This increasing separation raises the prospect

that there may be a need to invoke the safety net and

introduces a possible distortion into the international

payments system. As a consequence, both public-sector and

private markets have given great attention to attempting to

identify and control risk exposures. Perhaps one of the more

interesting developments in this evolution of regional and

globalized payments markets in both the public and private

sectors has been the push toward real-time gross settlement

systems with collateralization. Nowhere are these efforts

more apparent than in Europe, where the struggle to create a

single financial marketplace has focused attention and

generated analyses of the underlying issues, with the Bank for

International Settlements, the Group of Ten, and central

banks spearheading much of this work.

Casual empiricism suggests several reasons why the

systems are evolving in this direction despite considerable

analysis suggesting that netting arrangements are more

operationally efficient. The first reason is that systems,

instruments, and markets are evolving faster than the political

entities can bring their various rules and regulations into

harmony despite the many initiatives that have been

undertaken. Second, harmonizing systems to control

effectively the systemic risks (such as Herstatt risk) inherent

in nonsynchronized clearing and settlement systems, such as

foreign exchange markets, even if all the legal rules are in

place requires extensive international coordination

and cooperation. Third, central banks realize

that, regardless of the explicit rules governing

exchanges and settlement arrangements, they still may be

thrust into the role of the lender of last resort should major

participants get into financial difficulties that threaten to bring

down settlement and clearing systems. In the United States, the

decline in member bank reserve balances reduces payments

system participants' liquidity positions and increases the

likelihood that intraday credit may have to be extended.

Finally, the movement toward expanding the overlapping

hours that exchanges are open will increasingly make the

operation of net settlement systems more difficult.

(1.) See Benston and Kaufman (1995) for a review of the evidence on fragility

and systemic risk.

(2.) For a discussion of the risks and recent developments in exchange-traded

derivatives markets, see BIS (1997a).

(3.) Notice, however, that it may be a misnomer to call this type of event

risk, at least in the Frank Knight ([1921] 1971)

tradition,see also Hu (1994). The incidence of loss resulted from the

German governmental action, which seems almost

impossible to assign a probability to, and hence may be better

characterized as regulatory uncertainty. See BIS (1996) for a

comprehensive discussion of risks in foreign exchange markets and efforts

that both private- and public-sector entities have

made to identify, monitor, and control these risks.

(4.) Bankruptcy statutes can clearly affect the distribution of claim as

well. For example, some countries have what is known as a

zero-hour rule, which means that transactions taking place after the time

the institution is legally closed are regarded as invalid and will be unwound

(5.) See, for example, General Accounting Office (1994). An exception is

Edwards (1996), who describes the possible paths of a breakdown in

derivatives markets. He describes a scenario in which an end-user

fails to meet its obligations as a counterparty. This failure in turn

brings down a major dealer, thereby spilling over to both other

counterparties and dealers. These disruptions are then transmitted to

other markets as uncertainty both raises contract prices and leads to

reluctance to enter into contracts. There are then price breaks, credit

disruptions, falling asset prices, and, ultimately, real effects. Edwards

analyzes the likelihood that such a scenario would be realized and

concludes that true dealer credit exposures are small and substantially

smaller than their exposure on loans and other assets.


Bank of England and APACS. 1994. "The

Development of Real-Time Gross Settlement (RIGS) in the

United Kingdom." Bank for International Settlements

information release, April.

Bank for International Settlements (BIS).

1989. "Report on Netting Schemes." Prepared by the Group

of Experts on Payments Systems of Central Banks of the

Group of Ten Countries, February.

--. 1990. "Report of the Committee on Interbank

Netting Schemes of the Central Banks of the Group of Ten

Countries." November.

--. 1993a. "Central Bank Payment and Settlement

Services with Respect to Cross-Border and Multi-Currency

Transactions." Report prepared by the Committee on

Payment and Settlement Systems of the Central Banks of the

Group of Ten Countries, September.

--. 1993b. "Minimum Common Features for Domestic

Payments Systems." Report of the Committee of Governors

of the Central Banks of the Member States of the European

Economic Community. Action 2 of the report on issues of

common concern to EC central banks in the field of

payments systems, by the Working Group on EC Payment

Systems, November.

--. 1993c. Payment Systems in the Group of Ten

Countries. Report prepared by the Committee on Payment and

Settlement Systems of the Central Banks of the Group of

Ten Countries. Basle, December.

--. 1996. Settlement Risk in Foreign Exchange

Transactions. Report prepared by the Committee on

Payment and Settlement Systems of the Central Banks of

the Group of Ten Countries. Basle, March.

--. 1997a. Clearing Arrangements for Exchange-Traded

Derivatives. Report prepared by the Committee on Payment

and Settlement Systems of the Central Banks of the Group

of Ten Countries. Basle, March.

--. 1997b. Real-Time Gross Settlement Systems. Report

prepared by the Committee on Payment and Settlement

Systems of the Central Banks of the Group of Ten

Countries. Basle, March.

Benston, George J., and George G. Kaufman. 1995.

"Is the Banking and Payments System Fragile?" Journal of

Financial Services Research 9 (December): 209-40.

Corrigan, E. Gerald. 1990. "Perspectives on Payments

System Risk Reduction." In The US. Payments System:

Efficiency, Risk, and the Role of the Federal Reserve, edited

by David B. Humphrey. Proceedings of a Symposium on the

U.S. Payments System sponsored by the Federal Reserve Bank

of Richmond. Boston, Mass.: Kluwer Academic Publishers.

Edwards, Franklin R. 1996. The New Finance:

Regulation and Financial Stability. Washington, D.C.: AEI


Eisenbeis, Robert A. 1995. "Private-Sector Solutions to

Payments System Fragility." Journal of Financial

Services Research 9 (December): 327-49.

General Accounting Office. 1994. Financial

Derivatives: Actions Needed to Protect the Financial System.

Report to Congressional Requestors, GAO/GGD-94-133.

Washington, D.C.: Government Printing Office, May.

Hu, Jie. 1994. "Information Ambiguity: Recognizing Its Role

in Financial Markets." Federal Reserve Bank of Atlanta

Economic Review 79 (July/August): 11-21.

Knight, Frank H. [1921] 1971. Risk, Uncertainty, and

Profit. Reprint, Chicago: University of Chicago Press.


The author is senior vice president and director of

research at the Atlanta Fed. He thanks Clyde

Farnsworth, Craig Furfine, Diana Hancock, Pat

Parkinson, and Alice P White for helpful comments.
COPYRIGHT 1997 Federal Reserve Bank of Atlanta
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1997 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Eisenbeis, Robert A.
Publication:Economic Review (Atlanta, Ga.)
Date:Mar 1, 1997
Previous Article:Identifying monetary policy: a primer.
Next Article:Bank deposits and credit as sources of systemic risk.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |