The magic of an SDR-denominated bond

Stephen Paduano is a PhD candidate at the LSE and a former consultant for the United Nations Economic Commission for Africa. Brad Setser is a senior fellow at the Council on Foreign Relations and a former Treasury Department official.

Back in August 2021, the IMF agreed to a mammoth $650bn allocation of “Special Drawing Rights” (SDRs) — the ambitious but woefully underused reserve asset that countries can exchange for dollars, euros, pounds, yen, and yuan — to help combat the devastation caused by Covid-19.

The allocation was the chunkiest in the history of the SDR system — well beyond the $9.3bn allocation of 1970-1972, the $15bn allocation of 1979-1981, and the $250bn allocation of 2009. It also came with a commitment to fix the system’s biggest problem: most SDRs sit idle on the balance sheets of high-income countries.

Yet little progress has been made in piping surplus SDRs to where they were most needed. For a variety of reasons, only 11 countries have committed to rechanneling some portion of their SDRs, amounting to a mere $36.42bn. The US — the world’s largest holder of SDRs with $162bn — has been missing in action entirely. To understand the blockages, read this CGG paper by Mark Plant.

But with Western development budgets in decline, Chinese policy banks in retreat, and private capital unwilling and unable to fund the “big push” the global development agenda requires, we desperately need some creative thinking to find effective ways to mobilise these SDRs.

We reckon one approach warrants immediate consideration: World Bank issuance of a special series of SDR-denominated bonds.

SDR bonds could be held in a liquid form that satisfies the SDR’s requirement to remain a “reserve asset.” IMF members, particularly the US, would also find it legally and financially easier to purchase SDR bonds than to establish complex rechanneling facilities. And if the SDR bond pays the SDR interest rate, it would also offset the costs of using SDR.

The hows and whys and wherefores

Mobilising more SDRs requires solving three problems: SDRs need to remain a reserve asset on the balance sheet of countries contributing SDRs; SDR holders should not be put off by rechanneling costs; and the largest holder of SDRs (the US) needs to participate.

An SDR bond series issued by the World Bank — and possibly other development banks — would offer an elegant solution to all three challenges.

SDR bonds would qualify as a reserve asset. Many countries already hold Word Bank bonds in their reserve portfolios. Central banks could buy an SDR-denominated bond, earn interest on it, and trade it between each other and the world’s 15 other prescribed holders — thus ensuring its liquidity.

Furthermore, the bond could be SDR-denominated but cash-settled (with dollars or euros), which would simplify and streamline its use as a reserve asset. The World Bank has experience issuing SDR-denominated cash-settled bonds: it raised 500m SDRs ($675mn) in 2016 with a three-year fixed rate bond that paid a 0.5 per cent coupon in renminbi, just above the 0.25 per cent SDR rate at the time.

An SDR-denominated bond that pays the SDR interest rate also allows countries to rechannel their holdings at no budget cost. The interest rate burden that arises from holdings dropping below allocations would be offset perfectly by the interest income gained from the SDR bond

Finally, the US would also be able to buy an SDR-denominated bond and rechannel some of its $162bn holdings.

Its SDRs are governed by the 1968 Special Drawing Rights Act, which directed SDRs to be held by the US Treasury’s Exchange Stabilization Fund (ESF). The ESF statute permits the Secretary of the Treasury to “deal in gold, foreign exchange, and other instruments of credit and securities the Secretary considers necessary.” This means Congress has already authorised the US Treasury to purchase debt securities from highly rated issuers, which would naturally cover an SDR bond from the AAA-rated World Bank.

Engineering around the World Bank’s lending limits

The available “spare” SDRs would easily support an increase in the World Bank’s lending capacity by 50bn to 100bn SDRs ($60bn to $125bn) — a significant jump relative to its overall $425bn loan book. But the World Bank can only increase its net lending if funds raised from an SDR bond are not constrained by the current prudential limits on World Bank lending, or include an equity-like component.

A bit of background: The World Bank currently has around $50bn in capital — which backs $250bn in market borrowing and funds $240bn in IBRD loans. The Bank’s concessional lending arm is mostly funded by equity capital from accumulated grants, but it also now has a small amount of market borrowing (around $30bn). This means the World Bank’s existing lending is already backed by a significant equity base — with lending effectively limited by the rules limiting the Bank’s non-concessional lending to five times the Bank’s equity base.

There are calls for the World Bank to increase its lending simply by considering a portion of its large stock of callable capital as a form of equity. The new SDR bond could be viewed as a further direct commitment from the Bank’s leading shareholders to its financial model, with SDR bond purchases demonstrating political support akin to the current commitment of callable capital.

Pairing greater use of callable capital with the funds from SDR bonds thus could make such a change more palatable to the Bank’s board (and credit rating agencies). The new SDR bonds would be expected to be rolled over in the normal course of business, unlike traditional bonds, and thus should be viewed by rating agencies as a permanent contribution to the Bank’s resources.

But if additional capital is formally needed to preserve the Bank’s pristine credit rating as it expands its balance sheet, a portion of the new SDR bond issuance could be structured as subordinated debt.

The subordinated bonds would even pay a higher interest rate than the standard bonds — as long as the blended return on all the SDR instruments matches the SDR interest rate, there would be no net budget cost to current holders of SDRs who subscribed to both the junior and the senior debt. Some countries might also be able to use their existing SDR allocation to fund an expanded contribution to the World Bank’s capital base.

Just do it

The bulk of the last SDR allocation remains unused, buried under legal constraints and technical complexity.

Recycling the SDRs of both developed countries and over-reserved emerging economies into SDR bonds offers a way forward that preserves the reserve asset characteristics of SDRs, engages the US, and doesn’t affect budget resources. Legally and financially, it is far easier for most countries to buy development bank bonds than lend or give their current SDRs directly.

The world needs a bigger World Bank to boost financial flows to poorer countries, given declines in foreign aid and private financing, as well as for the longer-run and expensive clean energy transition. SDR bonds would be a simple and direct way to help mobilise the needed funds. Let’s do it.