The SDR as a Means of Payment A Comment on Coats
  • 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund
  • | 2 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund

Mr. COATS’s PAPER1 focuses on the development of a clearing arrangement for payments in private SDRs between individuals, between banks, and between countries. A network is conceived whereby a payment in private SDRs between two individuals in two different countries is ultimately settled by a transfer of “official” SDRs from one country’s account with the Fund’s SDR Department to the other country’s account at the Fund. A bookkeeping entry is taking care of the settlement, and the physical transfer of funds is thereby avoided.

Abstract

Mr. COATS’s PAPER1 focuses on the development of a clearing arrangement for payments in private SDRs between individuals, between banks, and between countries. A network is conceived whereby a payment in private SDRs between two individuals in two different countries is ultimately settled by a transfer of “official” SDRs from one country’s account with the Fund’s SDR Department to the other country’s account at the Fund. A bookkeeping entry is taking care of the settlement, and the physical transfer of funds is thereby avoided.

Mr. COATS’s PAPER1 focuses on the development of a clearing arrangement for payments in private SDRs between individuals, between banks, and between countries. A network is conceived whereby a payment in private SDRs between two individuals in two different countries is ultimately settled by a transfer of “official” SDRs from one country’s account with the Fund’s SDR Department to the other country’s account at the Fund. A bookkeeping entry is taking care of the settlement, and the physical transfer of funds is thereby avoided.

The implementation of the clearing system devised in the paper is commendable, as it would greatly facilitate the transferability of SDRs, but the question of the lender of last resort remains unanswered. Namely, Mr. Coats assumes implicitly that the sum of all the debits and credits flowing through the accounts of the various central banks that participate in the SDR clearing system at the International Monetary Fund would exactly match and that the clearing as a whole would show a zero balance. Unfortunately, as experienced in other clearing systems, the funds flowing to and from different accounts do not always balance, be it for technical or legal reasons. Clearing systems have established rules and regulations to cope with these shortfalls, and this comment tries to assess what could be done to alleviate shortfalls in the SDR clearing arrangement devised by Mr. Coats.

It is standard practice in international financial markets to advise the arrival of funds a couple of days before the settlement date. Normally, the recipient bank in country B, in turn, advises its client (B) of the arrival of the funds. This procedure greatly enhances the efficiency of the international payments system, because one can give payment instructions against cables of incoming funds. In other words, the recipient (B) of the SDRs may, in turn, transfer the expected SDRs to a third party (C) in a third country (C), effective the same value date as the expected receipt of funds. All other things being equal, the recipient central bank (B) has a neutral position on the payment date, because it receives an amount of SDRs from central bank A and pays the same amount to central bank C. But what if central bank A does not credit central bank B’s account on the payment date?

Such a mismatch will happen at the pinnacle of the clearing system whenever there is a disruption somewhere along the chain between individual, private bank, and central bank in the original paying country (country A). A disruption between individual A and his private bank should, in principle, not occur because the private bank will normally assure itself that the individual has adequate funds in its account, or a large enough line of credit available, before sending the advising cables. But a disruption between a private bank and its central bank is possible. For example, a payment by a private bank to its central bank to cover the use of its SDR clearing account may be delayed for technical reasons and may arrive too late to be settled on a particular day. Or, a private bank may erroneously believe that it has adequate funds in its clearing account with its central bank. Unless central bank A has made previous arrangements with the private banks participating in the SDR clearing account to cover such events, it will delay the transfer of funds to the recipient in country B. Consequently, the recipient central bank (B), instead of having a zero balance for that day, experiences a shortfall, as it has to pay the SDRs to the third country without receiving them from the original debtor’s country.

Central bank B could, in turn, delay the payment of the funds to the third country, but this could create a chain reaction that would disrupt the whole clearing arrangement. The main objective of the clearing arrangement being the maintenance of the integrity of the system, and hence the liquidity of the SDR, a backup procedure is needed. Several approaches are considered. First, the arrangement could provide that either the central bank that is about to cause a disruption (central bank A), or the central bank experiencing the shortfall (central bank B), could temporarily use its own holdings of SDRs with the Fund to fill the gap. If the latter applied, it is understood that central bank A is obliged to compensate central bank B diligently for the whole amount plus interest for late payment. A second solution is that the Fund lends SDRs from the General Resources Account to the central bank that is about to cause the shortfall (central bank A), or to the central bank experiencing the shortfall (central bank B), on condition that the SDRs be returned to the Fund when the shortfall is resolved. The central bank (A or B) advises the Fund that a shortfall is about to happen or has happened and asks if it can borrow the funds in order that the clearing arrangement not be disrupted.2

A third solution to alleviate the problem is that, at the inception of the clearing arrangement, all central banks wishing to participate in the system contribute a certain amount of SDRs, which are pooled as a buffer stock for possible disruptions. The SDRs could be either taken out of one member’s holdings with the Fund or purchased separately. To facilitate this procedure, the various central banks could ask the private banks that are holding SDR accounts with them in order to participate in the clearing arrangement to maintain a certain minimum balance. The interest earned when SDRs are borrowed from the pool could be earmarked to remunerate the holdings in the pool. If funds are needed over and above the total amount of SDRs available in the pool, one of the two other solutions must be envisaged.

The main goal of the clearing arrangement is the maintenance of the efficiency of the system and, consequently, includes the discouragement of disruptions. The three alternatives mentioned here have different implications toward meeting that objective. The first alternative—that the member country either causing or experiencing the shortfall temporarily decrease its holdings of SDRs with the Fund to fill the gap (and, for that matter, borrow SDRs from another member, if necessary)—has the advantage of avoiding the establishment of a formal backup system. But it possesses several disadvantages: first, this particular central bank’s holdings may happen to be inadequate; second, it possibly disrupts the desired level of SDR holdings by the monetary authorities of that country; third, and more fundamentally, if the central bank that did not cause the disruption is forced to take action and to incur costs in order to maintain the integrity of the clearing system, it cannot take retaliatory measures against the failing country, other than claiming reimbursement of all costs involved.

The second alternative—that the Fund extend a temporary credit to the central bank either causing or experiencing the shortfall—has the advantage of implying a supervisory role at the pinnacle of the clearing system. The possible drawback is that the extension of credit would have to be decided on a case-by-case basis, which could involve a certain degree of subjectivity. But a code of rules and regulations could alleviate this situation.

The third alternative—the creation of a pool of funds to finance possible shortfalls at the inception of the clearing arrangement—gives the same supervisory role to the Fund but eliminates the case-by-case creation of SDRs and the possible accompanying acrimonies. The pool could be created from a contribution of existing SDRs or a once-for-all allocation of new SDRs.

As mentioned earlier, alternatives two and three allow the Fund to administer the entire clearing system. To discourage failures, a penalty over and above the normal interest rate could be charged to the central bank that causes the disruption by not advising the Fund in time and by not providing financing either by using its own SDRs or asking the Fund for a bridging loan. And, if failures by one of the participants become recurrent, a procedure of exclusion from the clearing arrangement could be envisaged. One hopes that such stern measures, even if embodied in the clearing arrangement, would never have to be applied and that if an imbalance in the systems occurred, it would be purely accidental.

IN A RECENT PAPER, Coats1 argues that the direct use of the SDR—either issued by the International Monetary Fund or by commercial banks—in making payments is an important part of its development as a useful reserve asset. He also argues that the development of clearing arrangements is an essential element in the use of SDRs for making payments, and he describes one possible clearing mechanism, in which the SDR Department of the Fund has a key role to play.

The purpose of this comment is to point out the existence of one major issue that needs to be resolved if the widespread use of the SDR by the private sector—both as store of value and as means of payment—is to develop successfully: it is the issue of the determination of its price and the relationship of that variable to market equilibrium. Today, this variable is determined administratively through the formula “defining” the SDR. A similar procedure is used to determine the SDR’s interest rate. The problem is that, should a market for SDRs develop, there is no guarantee that those formulas will yield SDR exchange rates and interest rates that are consistent with market clearing.2 What is needed, therefore, is either a system in which the price of the SDR relative to all currencies—including that particular combination of currencies corresponding to the SDR “definition”—is free to adjust to clear the markets, or a system in which the supply of SDRs adjusts to the demand at the official price of one SDR for one basket of currencies. Given that there are presently no markets in which the price of the SDR can be determined, it appears necessary to maintain, at least temporarily, the administrative definition of the SDR. If, in addition, the private use of SDRs is to be encouraged, it is also necessary to organize a market in a way that guarantees that the market clears at prices that are consistent with this definition of the SDR. One mechanism that would yield this result could be the commitment of a group of central banks to sell or purchase any amount of SDRs for a corresponding number of “baskets” of the five relevant currencies. This group of central banks could be composed, for example, of the five central banks that issue the currencies used in the administrative definition of the SDR, and they could create a common “SDR Stabilization Fund” for the purpose of their intervention. Thus, for example, in Coats’s framework, should the commercial banks feel that they hold too much liquidity in the form of SDR-denominated deposits with their central bank, they could sell them on the foreign exchange market; the participating central banks would then purchase the excess supply of SDRs and issue, in exchange, deposits in the five underlying currencies. Through this mechanism, the stock of “high-powered SDRs” (i.e., the SDR-denominated liquid liabilities of the participating central banks) would be adjusted so as to clear the market. The scheme could be complemented by establishing a link between private and official (Fund) SDRs through an arrangement according to which participating central banks could adjust their asset position by selling to or purchasing from the Fund’s General Department any amount of SDRs necessary to maintain balance between their assets and liabilities in SDRs. This intervention mechanism could be discontinued later if the free market in SDRs grows sufficiently deep to allow for a meaningful determination of its price relative to national currencies.

To MAKE THE SDR a more important reserve asset, it must be made a more attractive asset for official holders. Progress has been made in this direction, but excessive attention has been paid to attributes of the SDR that bear upon its role as a store of value and too little attention paid to those that bear upon its role as a means of payment. Warren Coats sets out to rectify this error by offering a plan to make the private SDR an efficient means of payment.1 He is on the right track. But he does not stress sufficiently the most important contribution of his own proposal, and his scheme is unnecessarily radical.

If the official SDR were used to clear private payments, small but tricky differences in private SDRs would begin to disapppear. The official SDR would gradually become the standard SDR. This standardization must take place if the SDR is to be used eventually in foreign exchange trading, and that has to happen before the SDR can be used for official intervention.

More important, use of the official SDR to clear private payments would link it directly to the private SDR, and this linkage must take place to make the official SDR a useful reserve asset. Governments that borrow in SDRs should be able to add the proceeds to their SDR balances with the Fund and to use those balances to repay their debts. Governments that finance their balance of payments deficits by intervention in the foreign exchange market should be able to transfer SDRs to foreign exchange traders.

Under the plan proposed by Coats, members of the Fund would authorize their central banks to open SDR accounts for their own commercial banks. Such accounts could be used to clear transactions between banks in a single country. They could be used jointly with accounts at the Fund—official SDRs—to clear transactions between banks in different countries. Under this particular plan, however, governments would have to give up control over their own holdings of official SDRs. Whenever a French bank made an SDR payment to a British bank, official SDRs would be transferred automatically from the Bank of France to the Bank of England.

These and other difficulties can be overcome, however, by inserting a clearinghouse between central banks and commercial banks. Transactions between commercial banks would take place on the books of the clearinghouse, even those involving banks in different countries, and there would be no need for transfers of official SDRs. But transfers of official SDRs would take place whenever central banks (or governments) had dealings with commercial banks—whenever official institutions wanted to “transform” private SDRs into official SDRs or to go the other way—by borrowing, repaying debt, or intervening in the foreign exchange market.

Table 1 traces the transactions involved in setting up a clearinghouse. If Lloyds Bank wanted to join the clearinghouse, it would use its (sterling) balance at the Bank of England to buy official SDRs and pay them over to the clearinghouse. In this example, its subscription is SDR 100 million. The accounts of the Bank of England show a reduction of SDR 100 million in official SDRs held by the Bank of England and in its (sterling) deposit obligation to Lloyds Bank. The accounts of Lloyds Bank show the same reduction in the bank’s (sterling) balance at the Bank of England. It is offset by the bank’s SDR deposit with the clearinghouse. The accounts of the clearinghouse show an SDR deposit with the Fund and an SDR deposit obligation to Lloyds Bank.2

Table 1.

Setting Up the Clearinghouse

(Transactions in millions of SDRs and SDR equivalents of sterling)

article image

Table 2 traces a transaction between two commercial banks and shows that it has no effect on the SDR holdings of any central bank. In this example, Lloyds Bank uses SDR 75 million to buy yen from the Bank of Tokyo, and the SDR transfer is made on the books of the clearinghouse. (I omit the balance sheet of the Bank of Tokyo, as I do not need it to make my point.) The accounts of Lloyds Bank show its additional holdings of yen and the reduction in its SDR balance at the clearinghouse. The accounts of the clearinghouse show the SDR transfer from Lloyds Bank to the Bank of Tokyo.

Table 2.

An Interbank Transaction

(Transactions in millions of SDRs and SDR equivalents of yen)

article image

Table 3 traces the effects of intervention by the Bank of England. In this example, it uses sterling to purchase SDR 50 million from Lloyds Bank. (Lloyds Bank could be replaced by a foreign bank without changing the story in any significant way.) The accounts of the three institutions change in much the same way that they did in Table 1, but the signs of the entries are reversed. The Bank of England acquires SDRs from the clearinghouse, and they are official SDRs. Lloyds Bank acquires sterling from the Bank of England. The books of the clearinghouse reflect the “transformation” of private SDRs into official SDRs.3

Table 3.

Intervention by the Bank of England

(Transactions in millions of SDRs and SDR equivalents of sterling)

article image

The transactions shown in Table 3 are similar to those that would take place on account of borrowing by the Fund itself. The Fund would be able to issue debt to private institutions and take payment in official SDRs. If someone holding an SDR deposit with Lloyds Bank lent SDR 50 million to the Fund, the books of the bank would show reductions of SDR 50 million in its deposit liabilities and in its balance with the clearinghouse. The books of the clearinghouse would show what they do now—reductions of SDR 50 million in its deposit obligations and in its deposit with the Fund. The books of the Bank of England would not be affected. The General Resources Account of the Fund would show an increase of SDR 50 million in debt and an increase of SDR 50 million in holdings of SDRs (a claim on the SDR Department of the Fund). Until a clearinghouse is established, the Fund can issue debt denominated in SDRs but cannot collect the proceeds in official SDRs. It has to take payment in national currencies or in private SDRs.

Whenever a central bank or the Fund itself is involved in a transaction with a private institution, there is a change in the ownership of official SDRs. Transactions between private institutions, by contrast, affect the ownership of claims on the clearinghouse but do not affect the ownership of official SDRs.4

*

Participant in the Fund’s Economist Program.

*

Economist, Royal Bank of Canada.

The views expressed in this note are those of the author and not necessarily those of the Royal Bank of Canada.

*

Walker Professor of Economics and International Finance, and Director of the International Finance Section at Princeton University.

This comment has been adapted from a longer paper, “Use of the SDR to Supplement or Substitute for Other Means of Finance,” in International Money and Credit: The Policy Roles, ed. by George M. von Furstenberg, International Monetary Fund (Washington, 1983), pp. 327-60.

1

The SDR as a Means of Payment,” Staff Papers, Vol. 29 (September 1982), pp. 422-36.

2

A variation to this second solution is a temporary creation by the Fund of SDRs, which are lent to the central bank to cover the shortfall. Indeed, there is a temporary increase in world liquidity, but it is nullified when the SDRs are repaid to the Fund. The interest charged by the Fund for providing the SDRs is ultimately paid for by the central bank or private bank that caused the shortfall, so that (ultimately) SDRs are created in exchange for an equivalent amount of national currencies. But this option would require an amendment to the Fund’s Articles of Agreement.

1

Warren L. Coats, Jr., “The SDR as a Means of Payment,” Staff Papers, Vol. 29 (September 1982), pp. 422-36.

2

The analogy of gold may help to understand this point. Suppose that, when the official dollar price of gold was abolished, it had been replaced by an official price corresponding to a basket of the five main currencies of the world. This move would still have required intervention on the gold market to enforce the official average price.

1

Warren L. Coats, Jr., “The SDR as a Means of Payment,” Staff Papers, Vol. 29 (September 1982), pp. 422-36.

2

In consequence of the transactions shown in Table 1, the Bank of England is a net user of official SDRs and loses interest income; the clearinghouse is a net holder and earns interest income. In this particular example, the clearinghouse holds official SDRs and must therefore be given quasi-official status so as to qualify as a holder under Article XVII, Section 3 of the Fund’s Articles of Agreement. But other arrangements are easy to devise. The clearinghouse could be private but have an official sponsor, such as the Bank for International Settlements. Its sponsor would hold the official SDRs transferred to the clearinghouse; the clearinghouse would hold SDR certificates issued by its sponsor and backed fully by those holdings. Subsequent transactions involving official SDRs, such as the one in Table 3, would be handled by issuing or canceling certificates. I owe this suggestion to Jacques Polak, although he made it in a somewhat different context. (Note that there is no need to allocate new SDRs to the clearinghouse when SDRs are allocated to official holders. At some point, however, the commercial banks might have to make supplementary subscriptions, which means that central banks would have to make additional transfers of official SDRs. To this limited extent, official holders would still give up control over their SDR holdings.)

3

One reader of my first draft pointed out that the transactions in Tables 2 and 3, taken together, pose a problem for Lloyds Bank. It winds up with a debit balance in its account at the clearinghouse. Lloyds Bank would have to buy SDRs in the foreign exchange market or borrow them from other participating banks. The creation of a clearinghouse would probably give birth to an interbank market in SDR balances—the SDR counterpart of the Federal-funds market in the United States. (If all banks ran short of balances with the clearinghouse, because of large-scale official purchases, they would have to buy SDRs from their central banks to make supplementary subscriptions to the clearinghouse.)

4

Note that intervention by the Bank of England has the usual effect on the British money supply; the increase in the sterling balance held by Lloyds Bank constitutes an increase in the monetary base. If the effects of intervention are to be sterilized, it must be done deliberately. (Transactions between commercial banks, by contrast, do not affect the monetary base.)