Transforming Cross Border Payments

Cross border commerce and investments require cross border payments. While modern technologies continue to increase the speed and ease and lower the cost of domestic payments of domestic currencies, cross border payments remain relatively slow and costly. This article reviews the key features of domestic payments and their lessons for the transformation of cross border payments. The conclusion is that the natural convergence to a single currency as the vehicle between national currency payments across borders would be better served by an internationally issued vehicle/reserve currency settled on the books of the equivalent of an international central bank. Unanchored cryptocurrencies, such as bitcoin, are not serious contenders for this role.  The 189 member countries of the International Monetary Fund (IMF) have created the Special Drawing Right (SDR) for this role but the development of significant uses of the SDR unit of account in the private sector are required for the SDR to replace or at least supplement the US dollar in international finance and payments.

Domestic Payments

The considerable benefits of extending trade beyond family and nearby friends motivated the creation of ’money’ with which to pay for it.  Most money around the world is privately produced. In the United States, almost 90 per cent of broad money is privately produced (bank deposits). Thus, an understanding of how money and the means of paying it have evolved must focus on the private sector.  The evolution from cash payments (the hand to hand transfer of cash) to bank deposit transfers was slow. Not surprisingly it was driven by minimising the cost of making payments with the available technology.  Its history illuminates the challenges remaining for cross border payments.

Discussions of payments must clearly distinguish what is paid (medium of exchange) from how it is delivered (means of payment). The following stylised history initially focuses on the means of payment--the means of paying US dollars--and then considers alternative currencies. 

Metal coins, usually gold or silver, were a step up from barter but suffered a number of drawbacks that led to issuing paper certificates bestowing claims on more safely stored commodities. US dollar currency notes issued by banks in the US were redeemable for defined amounts of gold and were thus interchangeably accepted in payment of US dollar obligations in what we might now call peer to peer payments.  The evolution or establishment (as in the US Constitution) of a common unit of account (one dollar) enabled money to serve as the common invoicing/pricing unit so critical to economic efficiency.[i] 

All forms of money (banknotes, bank drafts, checks, etc.) ultimately reflected claims on gold or silver and were ultimately returned to their issuers who provided safe keeping services for gold and silver. The further a payment of cash was from the issuing bank the larger the discount tended to be, reflecting the higher cost of repatriating it and greater uncertainty of the issuer’s soundness. 

There were considerable advantages to paying large amounts by transferring ownership of deposits of cash (or gold) in banks via bank drafts (checks).  If the payment was to someone who also had a deposit in the same bank, the transfer was quick and simple (a debit to my account and credit to yours). Transferring a bank balance to someone at another bank was more complicated as it required shipping gold from the payer’s bank to the payee’s bank. To avoid having to ship cash or gold with each check, banks found it cheaper to open so- called correspondent accounts with each other and accumulate claims before actually shipping gold. My bank would credit my account with the amount of your check (my bank’s liability) and would credit its (correspondent) account of your bank (my bank’s asset).

Gold was only shipped between banks if a correspondent account balance got too large.  Payments and receipts by households, firms, cities, regions are roughly balanced over time.  Thus, to some extent the payments between customers of different banks are also roughly balanced. The balance is not between each pair of banks bilaterally but between each bank and all other banks. Check clearing houses developed to exploit this feature.

If each bank had to have a correspondent account with every other bank the number and related bookkeeping would be very large.  Banks quickly learned of the economies of dealing through a much smaller number of key banks in which they all kept their correspondent accounts. This is the now well-known hub- and- spoke arrangement also used by airlines to connect smaller cities with larger ones.  These hub correspondent banks evolved into local or regional clearing houses.  In each city all banks would gather in a room once a day to exchange the checks that their depositors had written between customers of different banks. The net amount for each bank against all others was recorded as a debit or credit to its account with the ’clearing house’ and only this net amount was settled by actually shipping cash or gold. 

To minimise the liquidity that banks needed to keep in their correspondent accounts, they began to borrow and lend clearing house balances between themselves (Interbank funds market). In all of these institutional developments the goal was to avoid the need to actually ship gold using the least possible resources and minimising other costs by the banks.

With the establishment of the Federal Reserve System in 1913, banks could replace their regional correspondent accounts with a single account with their Federal Reserve Bank through which good funds (i.e., those banked by gold) could be transferred to any other bank in the United States. This further simplified and reduced the cost of my paying you from my Washington, DC bank account to your San Francisco bank account without having to actually ship gold. The ultimate settlement asset became central bank money (a balance with the Federal Reserve).

The rules for intra bank settlements often lagged behind the technical possibilities. Physical checks had to be returned to the bank with the customer account on which it was drawn until 2004[ii] when banks were allowed to electronically return a photocopy of the check. In 1995, the Federal Reserve operated a fleet of 47 planes to fly checks to their originating bank at an annual cost of US$35 million (1995) dollars.[iii]  The fleet no longer exists.  Physical checks are increasingly being replaced by a variety of electronic payments both retail and wholesale (credit or debit cards, PayPal, Apple Pay, Pop Money, Zelle, etc.). These digital payments are exploiting (dependent on) the adoption of messaging standards, netting (as in ACH payments in which only net amounts are settled periodically such as daily) and increasingly user-friendly retail customer interfaces.

Innovation and competition for retail customers is intense but all of these payments are ultimately settled in central bank money (previously gold) and done so through banks. The two-tiered structure of a central bank, at which and through which all payments can be settled, and commercial banks as well as credit (Visa, American Express) and other payment service providers with deposits at banks, provides a good structure through which diverse payment approaches for the end users can compete. M-Pesa, the mobile phone payment system in Kenya, is an example of a very widely used payment vehicle for those without (direct) bank accounts.[iv]

Cross Border Payments

Cross border payments have lagged behind for several reasons. Generally, two currencies are involved requiring a market for exchanging them.  Each currency has one domestic central bank through which payments in that currency can be settled only between banks licensed in that country. Exchanging the currency paid for the one received must occur through banks with correspondent accounts from banks in each of the two countries involved. In order to offer currency exchange services these banks must maintain an inventory of each currency and incur the opportunity costs involved. A cross border payment often passes through a number of such accounts.[v]  The economies of a hub- and- spoke architecture, discussed earlier, have driven cross border payments to converge on a small number of intermediate so-called vehicle currencies to link the sender’s currency with the receiver’s currency. The US dollar dominates in this international reserve currency role, though American exploitation of the dollar’s dominance for political objectives (not always shared by other countries) is stimulating the search for alternatives.

The Digital Future

The gains already made and underway in making fast, safe, and cheap domestic payments can be achieved for cross border payments by applying the same technologies.  This requires an international rather than national currency. The member countries of the IMF have already adopted such a currency, the Special Drawing Rights (SDR), though its use has been grossly under-developed for lack of political will and some weaknesses in its design.[vi] Rather than allocating SDRs as now, the IMF should issue them according to currency board rules.[vii] Under currency board rules, the IMF would not have a monetary policy. It would sell or redeem its SDRs in response to the market’s demand at the fixed official price (exchange rate) of SDRs. This price would be given by the existing valuation basket of strong currencies or, in the future, by a basket of widely traded commodities.[viii]

If the supply of SDRs was issued according to currency board rules, it would transmit no monetary policy and would simply and easily satisfy the market’s demand for them at their fixed price. Thus, the French and German objections to Libra, the digital currency issued under currency board rules proposed by Facebook, undermining sovereign control of monetary policy seems misplaced. Market SDRs could be issued privately under currency board rules in the way banks now create private money (bank deposits).[ix][x] Market SDRs should be backed by, and redeemable for, “official” SDRs or assets of equivalent value (e.g. private SDR bonds). This would follow the so-called Chicago Plan or narrow banking model.[xi]

To the extent that SDRs replaced dollars as an international reserve asset, and for invoicing and settling internationally traded commodities such as oil, no exchange of currencies would be needed. Payments from anywhere in the world could be made to anywhere else in the world across the books of the international issuer in the same way that domestic currencies are settled across the books of their issuing central banks. Settlement could be instantaneous and final. While the SDR already denominates many official international obligations, such as those in the International Telecommunications Union and the International Postal Union (not to mention the IMF and World Bank themselves), it is not yet widely used in the private sector.[xii]

The improving private sector payment technologies can and should be applied to help expand the private uses of the SDR as a unit of account, medium of payment, and asset denomination.[xiii] Supplementing the IMF’s issuance of official SDRs with privately issued “market” SDRs would make a major contribution to those developments. SDRs would then be well positioned to replace or supplement the roles of US dollars and other national currencies in cross border payments.

[i] Warren Coats, “Free Banking in the Digital Age.” Banking & Finance Law Review. Vol. 33 Issue 3 p. 415 - 421

[ii] Check Clearing for the 21 Century Act enacted on October 28, 2003 by the 108th U.S. Congress.

[iii] Jerry Knight, "The check is in the airmail." Washington Post, January 17, 1996.

[iv] Warren Coats,  "The Technology of Money." Cayman Financial Review, January 18, 2012

[v] Committee on Payments and Market Infrastructures, Correspondent Banking, Bank for International Settlements, 2016

[vi]  Warren Coats,  "Real SDR Currency Board", Central Banking Journal XXII.2, 2011

[vii]  Warren Coats,  "The IMF should adopt a "real SDR'" Central Banking Vol XXX, No 1, Sept 2019

[viii] Warren Coats, "Returning to currencies with hard anchors" Adam Smith Institute, London April 15, 2019

[ix] Michael D. Bordo and Andrew T. Levin, “Central Bank Digital Currency and the Future of Monetary Policy.” NBER Working Paper No. 23711, 2017

[x] Tobias Adrian and Tommaso Mancini-Griffoli,  "From stablecoins to central bank digital currencies." IMF Blog, September 26, 2019.

[xi] A Guide to Banking Reform and Full-reserve (Narrow) Banking, http://www.narrowbanking.org/#top

[xii] George Hoguet and Soloman Tadesse. “The Role of SDR-denominated Securities in Official and Private Portfolios.” BIS Papers No 58, October 2011

[xiii] Warren Coats (2019). "Proposal for an IMF staff Executive Board paper on promoting market SDRs" Bretton Woods Committee blog Feb 19, 2019.

Dr. Warren Coats
Dr. Warren Coats is currently a fellow of Johns Hopkins Krieger School of Arts and Sciences, Institute of Applied Economics, Global Heath and the Study of Business enterprise who formerly worked at the International Monetary Fund as Chief of the Special Drawing Rights division and as director of the Cayman Islands Monetary Authority. He specialises in monetary policy. Warrens has won the Central Banking magazine 2019 Award for Outstanding Contribution for Capacity Building of central banks, especially in demanding circumstances. He holds a BA in Economics from UC Berkeley and a Ph.D. in Economics from the University of Chicago.