On April 17th a group of international banks called the Group of 20 gathered in New York to discuss their plans to create the Continuous Linked Settlement Bank, also called the CLS.
Among the G-20 banks are Bank of America, Chase Manhattan Bank, Citibank, J.P. Morgan, and Bankers Trust.
The CLS's mission seems promising enough: Eliminate the possibility that a default in the foreign exchange (forex) market will create a cascade of other defaults, leading to a global crash, by creating a clearing house that will settle all spot forex transactions worldwide.
But the CLS solution, by privatizing a big part of the global payments system and possibly affecting national monetary policies by influencing intra-day interest rates, may be too ambitious for the world's central banks to find comfortable, or even compatible with national sovereignty.
Such stakes, together with the fact that this story was written as the CLS met in New York, and during on-going negotiations between the G-20 and affected government organizations, may be one reason that most sources interviewed for this story insisted on speaking strictly on background; the stakes didn't, however, stop sources from expressing strong opinions. "The provision of payment and settlement services is in some sense analogous to garbage collection, and nobody gets uptight about privatizing garbage collection," says someone outside the G-20 with detailed knowledge of the CLS's plans. "I haven't seen anything which would indicate that there are any show- stoppers as yet, any central banks saying, 'Wait a minute, we don't think it's a good idea to have a CLS because they would be taking over things that we, the central banks, must do.'"
That's what you think, says Dr. Peter Garber, a Brown University professor of finance who co-authored a study of the issues and problems of foreign exchange settlement, including the CLS, for the Group of 30, a Washington, D.C.-based think tank. "It's a private global central bank," he says. "That's why there are major questions about jurisdiction and who would regulate it. The (central banks) definitely view (the CLS) as stepping on their toes, and have backed away from it."
What the G-20 itself has to say on the subject can't be related; it has been remarkably close-mouthed about its plans, say people familiar with the discussions-even in their conversations with central banks. Steven Thieke, the chairman of J.P. Morgan, who also serves as G-20 chairman, said through a spokesman that he was too busy to answer questions from the press.
Whether G-20 members are "garbage collectors," good citizens trying to save the world from economic ruin or powerful institutions bent on world domination, they're working in the world's biggest financial market. About 150,000 daily forex transactions take place 260 days a year, 24 hours a day, in a totally unregulated market, according to a 1996 International Monetary Fund (IMF) study. Depending on whether one or both legs of the deal is counted, these transactions are worth either $1.2 trillion or $2.4 trillion a day. Annualized, that's up to $624 trillion.
The world's securities markets volume, in comparison, totaled no more than three percent of that in 1995, according to the same IMF study- about $42 billion a day in a market plastered with regulations in most countries. Annualized, that's about $10.92 trillion.
The power of currency markets is legendary. Billionaire George Soros, for instance, is credited ad nauseam with driving down the British Pound in 1992, virtually by himself-thereby driving up Britain's trade deficit. A crash in a market that important could hardly escape notice, even on Main Street.
It follows that putting seat belts on a market like that is a good idea. And in fact, the G-20 fielded its proposal for the CLS last March, about the same time that the Committee on Payment and Settlement Systems (CPSS) of the G-10 Countries, meeting at the Bank of International Settlements (BIS) in Basle, Switzerland, put its gun on the table by saying in its Allsop Report that either the private sector find a solution to the problem in two years, or the CPSS's members would do it for them. Other recommendations in the Report, commonly called the Orange Book, include a call for national real time gross settlement (RTGS) systems and better risk management programs at banks.
The premise behind the Orange Book's recommendations is that while it's possible to eliminate so-called Herstatt risk-named after the 1974 collapse of Frankfurt-based Herstatt Bank, which caused a cascade of forex- related defaults in the world's currency markets-only some steps to do so have been taken by the industry.
Mitigating Risk in the Forex Market
As recently as 1995, for instance, a cascade was narrowly averted during the Barings Bank collapse when a party making a forex payment to one of Barings's correspondent banks tried to back out of the deal. The problem was resolved with a loan and avoided a disaster that would have involved 45 other banks. Other cascades were narrowly avoided during the invasion of Kuwait, the Russian coup, the collapse of the Bank of Commerce and Credit International (BCCI), and that of Drexel Burnham Lambert.
Despite such scares, life in the forex market goes on, if only because there's so much money to be made: One forex consultant for instance, Barbara Rockefeller, claims that following her advice from 1990 to 1996 would have realized an unleveraged, uncompounded, 9.05 percent annual return. Another good reason: Global trade requires hedging to damp currency risk.
As a result, efforts to mitigate Herstatt risk are already in place, CLS or no. In 1993, for instance, the Bank of Japan lengthened its settlement hours to 5 PM local time; and beginning this year, the U.S. Federal Reserve System's FedWire payments system will be open 18 hours a day. Before the FedWire action, the Bank of Japan closed before FedWire opened, creating a lag during which anything could happen. The result of this seemingly unspectacular adjustment is that all central bank settlement systems are open during overlapping hours, leaving fewer forex deals hanging fire for settlement.
RTGS systems, which are heavily computerized, are also spreading rapidly. France is building one, and Germany is about to as well. In Australia, Tandem Computer Co. is building an RTGS system that clears both currency and securities transactions. And the national RTGS systems in Europe are working together to create TARGET, or Trans-European Automated Real-Time Gross Settlement Express Transfer, which would link them.
Eliminating Lag Times
RTGS systems are popular with the CPSS because they provide a technical platform for eliminating settlement risk. In an RTGS system-like FedWire, for instance-payments are settled on a pay-versus-pay (PVP) system. All the money changes hands at once and the deal is final. In end- of-day systems, like those still operating in many places, settlements can be delayed for as long as five days. Since banks doing a lot of forex sometimes risk their entire balance sheets on one deal, a bad guess could easily be fatal; and since central banks would have to step in during a crisis as lender of last resort, lag times like that are dangerous because they allow time for other events to happen.
While RTGS systems could theoretically eliminate Herstatt risk, they require arrangements-like a common method of verifying that payments are actually made-that don't yet exist. Also, according to the IMF study, linking all RTGS systems worldwide into a big "system of systems" would mean that problems in one country would temporarily affect the whole system.
The pitfalls and challenges of smoothing out the crimps in systems like this are visible in Europe's faltering momentum towards making the Euro an actual currency after 50 years' negotiations.
While the the world of forex gropes for an ideal solution, a patchwork settlement system is in place that central bankers find imperfect, but manageable, and which doesn't challenge their power.
Actual settlements are done by central banks, after passing through clearing houses or netting schemes. Settlement can be either through real time gross settlement systems like FedWire, or by end-of-day settlement systems like Germany's.
Clearinghouses, like New York's Clearing House for International Payments System (CHIPS), and Clearing House Automated Payments System (CHAPS), based in London, routinely clear billions of dollars a day in forex payments; CHIPS alone, which clears about 70 percent of all U.S. dollar deals, processes some $600 billion in forex transactions a day, according to George Thomas, a CHIPS svp, who adds that, pending legal decisions, CHIPS is about to launch an RTGS-like system of its own. "We're going to say, if you fail, you pay," he says. "We have it in principle now, and we've tested it. We expect to make a decision to put it in place by the end of the year."
Beside the clearing houses, groups like the Society for Worldwide Interbank Financial Telecommunications (S.W.I.F.T.), Citicorp Dealing Resources-operated FXNET, Exchange Clearing House (ECHO) and Multinet operate so-called bi-lateral and multi-lateral netting schemes, in which only the difference between transaction "legs" is exchanged. By reducing the amounts exchanged, risks are reduced. FXNET and Multinet recently formed an unspecified alliance.
Creating a Global PVP System
If the CLS becomes reality, it will replace some of these institutions and subsume others. What's envisioned, according to a source within the G-20, is a bank that will handle simultaneous booked transfers between counterparties. Each counterparty would have a single account at the CLS with sub-accounts for each currency in which they trade, and settlements between the parties would be made on a multi-currency, intra- day basis-although not in real time-effectively creating a global PVP settlements system for spot transactions.
RTGS systems play a key role in the CLS solution, according to the outside expert with knowledge of the G-20. "You need a domestic RTGS system in order for the currency to participate in the CLS," he says. "The actual money moving in and out of the CLS has to come in in final form, as settled money from an RTGS system. Those who are interested in (forward trade) pre- settlement risk mitigation could use a MultiNet or ECHO, and that could feed right into the CLS on the settlement side; those who are not interested in pre-settlement risk could feed right into the CLS."
The main G-20 banks are already making sure they don't miss anything, should their plans jell. Most of the partners in FXNET, for instance, are also G-20 banks, and a number are also partners in ECHO; three are partners in Multinet.
The G-20's plans might seem tidy. But in some ways, they also make central banks subsets of its operations-an unappealing prospect to your average central banker. A central bank, for instance, "might feel it had to intervene to bail out the system if excessive credit has been given to a bank in its country, and there's a fail on it. Then, there would have to be an acquisition of collateral from whoever is being bailed out, and (the loan) would be uncollateralized," says Brown University's Garber.
This what-if scenario raises the issue of whether the injection from the central bank was going to be sterilized-repaid before the close of the business day-or if it will persist overnight. "Then you get into the impact on monetary policy because overnight rates will start to get affected," Garber says.
The 1996 IMF study raises another possibility; collateral in the form of government bills would be needed to back CLS-destined forex deals. This collateral would have to stay in the vaults of the banks in question and wouldn't be available to the central banks for repurchase purposes. This in turn would crimp the ability of affected central banks to manage intra-day credit.
Since the CLS would more likely be located in London than New York because London straddles the time zones of the world's currency markets, this issue is more one for the Bank of England to worry about than for the Federal Reserve. But since The Bank of England is now trying to establish a liquid repo market in Gilts, it's unlikely to greet obstacles to that effort with any great enthusiasm.
Looked at this way, eliminating Herstatt risk by allowing the CLS to open for business may be looking too pricey for the central banks. This is especially true because the system now in place is made of relatively small organizations, which favors them politically.
The CLS, on the other hand, which would be composed of the major forex banks clearing each other's trades, would represent a substantial counterweight to both their domestic and international powers.
And since the G-20's plans will get nowhere unless every central bank affected gives them their blessing, it means that the CLS's most vocal critic-concerned, perhaps, for its national grandeur-has a veto on its future. This doesn't even factor in the prospect of domestic politicians posturing about the CLS on the international stage for domestic consumption. The G-20 could easily find such forces to be, at the least, irritating obstacles. At best, this can only mean a long road for the G-20 to travel, during which it will have to maintain political cohesion to salvage any chance of success.
Cool Reception to Privatization
The central banks are clearly not warming to what they've seen so far about the CLS. "It's been a year since the (Orange Book) came out, and I know of nothing in the discussions that have convinced the central banks to change their stand that the best solutions, to the extent they can be created, should come from the private sector," says Lawrence M. Sweet, a vp at the Federal Reserve Bank of New York who recently rejoined the bank after being seconded for two years to the CPSS. "Nor am I aware of any change in the view that it's for the market-not central banks-to determine the optimal number of complementary and competing services. It's not clear if the market might be better served by either a small or a large number of service providers; central banks have not endorsed any one service as a superior solution."
This perspective can only delay an effort that has already been set back by the G-20 from a 1999 launch date, to as late as 2001. And it may take even longer. "The issues surrounding creating the CLS are technically difficult, and all the central banks will have to think hard about them,"says one of the co-authors of the IMF, voicing her personal opinion. "I think it will be quite some time before the CLS is up and running."