How an SDR Denominated Bond Could Work

This is a joint post by Brad Setser and Stephen Paduano, the executive director of the LSE Economic Diplomacy Commission and a PhD candidate at the London School of Economics. The two proposed that the World Bank should issue an SDR denominated bond to mobilize funds to expand the World Bank's balance sheet.  According to the reporting of Shabtai Gold of Devex, the idea is now under active consideration. Gold reports: "The idea is complicated, but "nothing is impossible,” a World Bank source told Devex regarding the proposal. It could get done if the big shareholders want it to".

This blog answers some of the most common questions about how such a bond could work.  Many of the questions were originally posed by Devex.

Why do we need an SDR denominated bond? 

A number of countries are sitting on large stocks of SDRs—the IMF reserve asset that can be exchanged for dollars, euros, pounds, yen, and yuan. They want the capacity to maintain those SDRs as part of their reserves. They have difficulty giving those SDRs away, but can invest their SDRs in classic reserve assets—like bonds. An SDR linked bond is in fact a perfect match for these countries, as they technically have to pay the SDR interest back to the IMF on any of their SDRs that they use.

The World Bank is an institution that raises money mostly by issuing bonds, and then uses the funds it raises to make loans to low- and middle-income countries. Right now, there is a shortage of reasonably priced private financing for the frontier markets, and China policy lenders have also scaled back. An SDR denominated bond would help the World Bank to expand its lending safely, as it would allow the bank to tap what is, in effect, a captive market. There would be an expectation that the SDR bonds would roll in perpetuity, and thus the Bank would essentially have a permanent source of financing. A bigger Bank would allow more financing of green infrastructure, without any reduction in the Bank’s current activities.

An SDR bond also would potentially help the United States make use of its $160 billion in SDRs. Congress has already granted the U.S. Treasury the power to buy securities from highly rated issuers, like the World Bank.

How exactly would an SDR bond at the World Bank work?

The World Bank can issue a bond denominated in SDRs in the same way it issues bonds denominated in dollars, euros and other currencies. Countries that currently have surplus SDRs on deposit at the IMF would lend those SDRs to the World Bank, which would then convert them into usable currencies through the IMF. The countries that lent their SDRs to the IMF would receive an SDR-linked bond in exchange. We suggest that the bond could settle in dollars or euros, which would make the bond easier to trade—and thus, help these bonds easily qualify as a reserve asset.  The World Bank has already issued a small SDR-linked bond which settles in China’s currency (the yuan), so such bonds don’t require any significant new innovations.

In order for such bonds to raise the World Bank’s overall lending, they need to be a part of an overall plan to expand the World Bank’s overall balance sheet—an SDR-linked bond shouldn’t just substitute for the Bank’s current issuance in dollars, euros and other currencies. The SDR bond is a particularly low risk instrument for the World Bank, as they would be issued to the Bank’s main shareholders and there would be every expectation that the stronger shareholders would never actually need to sell their bonds into the market to raise cash.  Thus, these bonds should naturally rollover at par when they mature. Therefore, they provide a particularly attractive means to stretch the World Bank’s existing capital a bit more, as the G-20 working group on the multilateral development banks’ capital adequacy framework proposed.  

An SDR bond could also be paired with a broader reform and funding package that adds to the Bank’s working capital through the issuance of subordinated debt or other forms of hybrid capital, or even through a combination of a small general capital increase and an increase in the Bank’s internal leverage limit that stretched all of the Bank’s capital a bit more. Some countries may even be able to use their existing SDRs to buy a hybrid capital instrument. The basic idea is to make use of the pool of SDRs currently sitting on the balance sheets of the Bank’s main shareholders and provide a source of cheap, reliable, and safe funding for a bigger World Bank.

Would big shareholders need to sign up in advance? Would Treasury need Congress to approve such an SDR purchase?

The big shareholders would need to commit to buying the bond at par (so they receive exactly the SDR interest rate on their bonds) as part of a broad commitment to expand the World Bank’s lending capacity. They would not need to commit to hold on to the bonds—after all, the bonds should be designed to be a reserve asset that could be sold to another shareholder of the Bank or into the private market for cash. But the idea is that the bonds would in effect be placed directly with the Bank’s shareholders.

The U.S. Treasury would likely want to consult Congress before subscribing to such a bond, just to avoid any misunderstanding. But  the statute governing the Treasury’s Exchange Stabilization Fund already provides the U.S. Treasury with the legal authority to buy a World Bank bond, as the Exchange Stabilization Fund is clearly allowed to deal in securities. Congress has given the Exchange Stabilization Fund that power for good reason! It would be a hassle for all parties involved—Congress, Treasury, and the issuer of a bond—to go through a lengthy process just to buy and sell a safe security.  That is something that the Exchange Stabilization Fund, or any other holder of foreign exchange reserves, does routinely.

The ECB has expressed concerns about rechanneling SDRs to multilateral development banks. Would the European countries be able to buy an SDR bond?

We believe that most European countries would be able to subscribe to a well-designed SDR bond issue. ECB President Lagarde has noted, in the context of the discussion of other proposals to make use of the world’s spare SDRs, that “retaining the reserve asset status of the resulting claims is paramount.”  A World Bank SDR-linked bond would explicitly be designed to qualify a reserve asset, and to create an SDR-linked claim that is “highly liquid and of high credit quality.” The ECB also expressed concern about rechanneling SDRs though the MDBs rather than the IMF. Our proposal though is different, as we would use the SDRs for direct balance sheet financing—the funds from an SDR bond would be combined with funds raised through the World Bank’s current market borrowing to support its overall activities. 

There equally should be ways to structure an SDR bond so that it does not raise concerns about the use of the Eurosystem’s reserve assets for monetary financing. Conceptually, the purchase of a SDR denominated bond is no different from the conversion of existing SDRs into dollars by one of the member central banks of the Eurosystem and the purchase of an existing World Bank bond with those dollars. Such a chain of transactions would clearly not impact the single monetary policy of the Eurosystem, help finance budget deficits inside the euro area, or change monetary conditions in the euro area.

We consequently believe the ECB could simply exempt an SDR bond issued by the World Bank from the prohibition on monetary financing, as it has already done for the IMF’s Poverty Reduction and Growth Trust (PRGT) and the Resilience and Sustainability Trust (RST). The issues involved with a World Bank SDR bonds are less complex than those posed by the AfDB and IDB’s hybrid capital proposal.

There are also other possible paths forward. Euro area countries could buy SDR bonds in the secondary market—perhaps directly from countries like the United States and the United Kingdom. Such purchases would fall within the existing Eurosystem's existing practice of purchasing bonds, including bonds issued by European supranational institutions, in the secondary market. Alternatively, if euro area countries consider themselves unnecessarily constrained by ECB policy, they can adjust their SDR holdings to map the practices of other countries, such as the US, which holds its SDRs in the Treasury not the Federal Reserve. In many countries, both the fiscal authorities and monetary authorities hold international reserves, and the IMF considers the member country (not its central bank) the prescribed holder of SDRs. No specific ECB policy that would prevent Eurosystem national central banks from reapportioning or swapping reserves like-for-like with fiscal authorities.

How many SDRs are out there that could be put to use? What size should the bond offering be? 

There are $935 billion worth of SDRs in the world today. A small fraction of these—no more than $45 billion—have already been committed to support the IMF’s Resilience and Sustainability Trust. A bigger fraction is held by countries that are themselves short on reserves and therefore wouldn’t necessarily want to invest in an SDR-denominated bond. Several hundred billions of SDRs are held by the G-10 countries and China—countries that aren’t likely to ever need to directly exchange their SDRs for cash.  

The exact size of the bond issue—or more realistically, a series of bond issues—would likely be limited by the World Bank’s capital framework. But we hope that the World Bank would think ambitiously. The EU, the UK, the US, Japan, and China collectively received over $300 billion of SDRs in the most recent SDR allocation—a commitment by the Bank to say double the size of the IBRD (the Bank’s non-concessional loan) could, for example, be fully funded through a set of SDR-linked bonds.  

Could the bond issuances be done by the usual Wall Street firms?

An SDR bond ideally would be placed directly with the Bank’s main shareholders, so the role of Wall Street firms would be minimal. There would be no need for book-building and the like. The only real requirement is that the bond be structured to trade commercially if needed. The fact that the bonds can be exchanged in the private market easily and efficiently would help to clearly establish that they qualify as liquid reserve assets. Some Wall Street firms might be willing to commit to establish and to maintain the associated infrastructure as part of their broader commitment to good global citizenship. The existing precedent is a bit unique: the World Bank’s first SDR bond was issued in China and designed to appeal to “private” investors—and in that case, the lead managers were the large Chinese state commercial banks and HSBC.   

Who will be able to access these bonds?

While SDRs can only be held by “prescribed holders,” which are the 190 member countries of the IMF plus 15 other entities (such as multilateral development banks), an SDR-linked bond could in principle be held by any private investor (including Wall Street banks' clients, such as pension funds). Once the SDR bond is issued, in theory, it could trade freely. The bond would be designed though to be held at par by the world’s big existing holders of SDRs, as they would receive an SDR-linked bond with an interest rate that matches perfectly the interest rate on their SDR denominated liability to the IMF’s SDR department (when SDRs are created, the IMF’s shareholders effectively borrow SDRs from the IMF in perpetuity, and in return receive a deposit of SDRs in their SDR account at the IMF—so SDRs aren’t actually free money). Because the bonds would be designed to help the large existing holders of SDRs match their assets and liabilities, we would expect that the bulk of the bonds would remain on the balance sheets of the world’s biggest holders of reserves.

Why has this not been done before? 

The main reason why this hasn’t been done before is that there haven’t been that many SDRs out there—but that changed with the most recent SDR allocation. 

There will certainly be pushback from people who simply don’t like change. And some people who opposed the issuance of SDRs probably would prefer that the bulk of the last SDR issuance remain unused and idle.

The strongest objection is that the World Bank doesn’t need more funding, as it has no difficulty raising dollars and euros by issuing bonds today.  But the point here is that by raising money directly from the Bank’s biggest shareholders—the same countries that back the bank’s existing bonds with their commitment of callable capital—the Bank can safely stretch its balance sheet and leverage its equity base without putting its rock-solid rating at risk.

What would the capital raised by the bonds be used for?

It would be available to fund all the activities of the World Bank Group. At the moment, there is a particular need to meet clean energy transition targets and to help finance fragile and conflict prone states. But the World Bank can do anything it wants with these funds.  

There would be some additional complexities using the Funds for fully concessional loans—as a donor would likely need to commit to cover the interest rate costs. But at the end of the day, that is a surmountable problem with a bit of political will, as it would be a good way to really stretch a country’s aid budget.

What are the chances of this idea becoming reality?

Better than most think! Maybe fifty-fifty, though neither of us is a professional forecaster. We can say this: the proposal is ready-to-go in terms of how the World Bank would issue it, how large SDR holders such as the United States would buy it, and how the SDR bond itself would work. There are no serious technical obstacles standing in the way. If senior officials at the World Bank and the large SDR holders would like an SDR bond, they can make it a reality relatively quickly.