Tokenised SDR

UBS and RBWC Conference-Managing the soft landing of the global economy

Session—50th anniversary of the SDR: A new life?

Wasington, D.C., 12 April 2019

Ladies and Gentlemen,

I'm delighted to speak here today and am most grateful to the organisers for the invitation. In my brief remarks, I will focus on an idea that was advanced by IMF Managing Director Christine Lagarde in 2017, namely to offer a digital SDR. Yesterday, at the IMF Seminar Money and Payments in the Digital Age, Lagarde asked if a digital SDR would be possible and one discussant responded that it could become functional within 12 to 24 months. At Accenture, I'm working on a number of projects and proposals that involve digitalising or tokenising central bank money. Hence, a tokenised SDR seems very much in line with an incipient new openness to give considerations to tokenising national currencies. But given the difficulties of amending the existing SDR, a new approach is needed.

The innovation would not be for the SDR to be digital. It is already a book currency. The innovation would be to tokenise it and issue it outside the IMF. Let's call it Tokenised SDR or TSDR. The idea would be that central banks offer TSDRs using their IMF SDRs as a hedge against obligations arising from TSDRs so to effectively pass-through the SDR to the market similar to an exchange rate target. It would allow to mobilise the SDR that has remained largely dormant—few central banks if any boast that they hold SDRs and many perceive them as a burden—as a financial resource. Equally importantly, it would not require any changes to the existing provisions governing allocations and operations of the SDR.

The use case for TSDRs is to serve as a medium of exchange in token-based ecosystems to enable settlement in official currencies. Those ecosystems are now emerging and include but are not limited to digitalised exchanges, regional local currency payments integration, trade finance and value chain management. The main international currencies are not yet established in those ecosystems. As an official asset, the TSDR would thus enjoy a fist mover advantage. TSDRs could bring certainty and confidence to those ecosystems, be used as the cash leg in token exchanges, be a hedge in international payments and offer a stable unit of account. For regional blockchain-enabled payment systems, TSDRs could be used as a vehicle currency to invoice and conduct payment transactions or as collateral in local currency operations.

The SDR is 50 years old. Since its inception though in 1969, it seems fair to say that the IMF has been less than enthusiastic about the SDR. It never made it to become the primary international reserve asset as intended and stipulated in the IMF Articles of Agreement. In a 2018 discussion about the role of the SDR, the IMF Executive Board concluded that most Directors “were uncertain or unconvinced that there is a role for the SDR […].” The failed success of the SDR can be attributed in my view, and others have similarly argued, in large part to a faulty design. A 2018 IMF Policy Paper about the SDR resolved that the SDR could play a greater role but that significant changes would have to occur including for its allocation mechanism and legal framework. Allocations of and changes to the SDR require a large qualified majority at the IMF Executive Board and may involve amending the IMF Articles of Agreement which has proven very difficult in the past and for which sufficient consensus is not likely to emerge any time soon. For that reason there have only been three general SDR allocations, issuance of SDRs, and one special allocation since inception. Today there are US$290 billion SDRs outstanding.

SDRs are only available to member countries of the IMF and can only be used to settle transactions within the IMF and a small number of designated agencies. Just to clarify, the SDR is not a currency. Nor is it a liability of the IMF. It is a potential claim on the feely usable currencies of IMF member countries. In principle, it represents an option that can be put to a member country at any time to obtain a freely usable currency at a price equivalent to the SDR basket exchange rate. The SDR basket is simply the weighted exchange rates of the SDR basket constituent currencies, the dollar, yen, sterling, euro and renminbi, against the U.S. dollar. In that sense, the SDR can be loosely considered a currency basket or an international currency or an international reserve asset. But it really is a unique asset.

TSDRs would allow central banks to manage more flexibly their SDR exposure. With an SDR allocation, countries receive an asset in SDR and also a liability in SDR. Any difference between allocation and holdings, as SDRs can be freely bought from and sold to other countries, results in a net SDR exposure. Naturally, the approach implies that the total amount of TSDRs may not exceed existing SDRs. Also no new financial resources would be created.

TSDRs would be a liability of a central bank and created by issuing an asset denominated in SDR on a blockchain maintained by a permissioned network operated by the IMF member countries. TSDRs could include smart contracts, self-executing transactions triggered by a pre-determined event, that would ensure convertibility. The blockchain-based approach would allow TSDRs to be transferred in peer-to-peer transactions that would not involve the issuing central bank.

Market participants would purchase TSDRs for say dollars from issuing central banks at par on a digital exchange. The TSDR price would be equivalent to the SDR basket. Holders would be able to exchange TSDRs freely and TSDRs could be put back unconditionally to the issuing central bank or any participating central bank. In the event of an appreciation of the SDR against the dollar, the central bank could sell its SDRs to another SDR holder and obtain the needed dollars for redemption. In the event of a depreciation of the SDR against the dollar, the central bank could choose to simply redeem the lower dollar amount. The movement in and out of TSDRs should be seamless. TSDRs should be detokenized and tokenised flexibly on demand.

There are of course private SDR-type digital currencies in the market, like for example the XDR coin or the ACC coin but to date with very limited if any success. Stable coins, digital coins that are backed by national currencies similar to a currency board, are trying to offer the need for a stable unit of account in tokenised ecosystems. However, those private initiatives struggle to attract the confidence needed to be used widely.

The IMF could encourage a common contract standard for TSDRs to enable TSDRs to be fully fungible among participating central banks. Central banks may have to meet certain criteria to join a TSDR scheme to ensure needed trust in TSDRs.

TSDRs could be issued now. They would help mobilise existing SDR resources for markets and offer a new approach to international payments. Few doubts that tokenisation of financial assets and non-financial assets will not play a significant role in market developments over the medium term. A TSDR may just be the right medium of exchange to give confidence that such developments rest on solid multilateral foundations of which the TSDR would be one of its most innovative exponents.