The global debt architecture rewards the wrong creditors

THE GLOBAL debt architecture was supposed to ensure that sovereign defaulters faced their creditors with equity. Instead, Ethiopia has demonstrated that private bondholders with London court lawyers extract better deals than taxpayer-funded governments.

On June 29th Ethiopia announced a preliminary agreement to restructure its $1 billion Eurobond, on which it defaulted in December 2023. The deal cuts the bond's principal by 12%, replaces it with a new $880 million instrument maturing in July 2029 at a 6.15% coupon, and pays bondholders nearly $155 million in past-due interest. It also attaches a warrant giving holders the right to subscribe to a future international bond of up to $1 billion. Private creditors representing about 45% of the notes have agreed in principle. The IMF and the official-creditor committee have offered non-objection. The machinery seems to be turning at last.

The trouble is what the deal reveals about the system that produced it. Under the G20 Common Framework — the multilateral architecture created in 2020 to handle the debt of poor countries with official lenders — Ethiopia's bilateral creditors, co-chaired by China and France, accepted a present-value loss of 12.5% on their debt. By one independent calculation, the private bond restructuring offers bondholders a net present value of 105.9% of what they are owed, versus 82.7% for official creditors. Private investors are set to recover more, in present-value terms, than the governments that lent to Addis Ababa.

That inversion is not an accident. It is built into the structure of the Common Framework, which was designed to coordinate official creditors but has no direct authority over private ones. Private bondholders are governed by contract law and the threat of litigation. The Framework's "comparability of treatment" principle — meant to ensure all creditors share the burden equally — exists only as a negotiating norm, not a legal requirement. It works only when private creditors lack leverage and official creditors hold the purse strings. Ethiopia's case shows the reverse.

The private deal collapsed once before, in January 2026, when official creditors rejected the terms as too generous. Bondholders then threatened legal action in the United Kingdom. Faced with litigation risk and a stalled IMF programme, Ethiopia offered a new proposal in May. It, too, was rejected. The June agreement splits the difference: a modest nominal haircut, full payment of arrears, a warrant that provides the present-value sacrifice the official committee needed, and a future bond option that gives private investors upside if Ethiopia's creditworthiness recovers. It is a deal that satisfies the letter of comparability while violating its spirit.

To be sure, the numbers are not straightforward. Official loans carry concessional terms — lower coupons, longer tenors — that make a direct comparison with a commercial bond misleading. The bilateral rescheduling runs until 2038-39 and lowers debt-service payments by 34% during the IMF programme period, without reducing nominal principal. Private bondholders, by contrast, accept an immediate face-value cut and a shorter maturity. The warrant's value depends on Ethiopia's future borrowing costs, not on export revenues as in earlier proposals, which sidesteps the risk of government manipulation. On a cash-flow basis during the programme period, private creditors actually receive less than their official counterparts. The warrant is what tips the present-value calculation in their favour.

Yet the principle remains. Private creditors with legal muscle — including VR Capital and Farallon Capital Management — extracted a deal more favourable in present-value terms than the one their official counterparts, whose funds come from taxpayers, were prepared to accept. That outcome would be defensible if the system rewarded market discipline. But the system does not. It rewards litigation capacity and the opacity of bilateral negotiations, which shield official creditors from public scrutiny while leaving private bondholders to litigate in open courts.

The broader architecture is equally dismaying. Ethiopia first requested debt treatment under the Common Framework in early 2021. Five-and-a-half years later, the process is still incomplete. As of mid-2026, bilateral agreements have been signed with France and Italy; China, Ethiopia's largest bilateral lender, has yet to finalise its deal. The framework operates creditor by creditor, committee by committee, with no collective decision mechanism. Each creditor delays to extract a better position; the debtor accumulates arrears in the meantime. The IMF's fourth review of Ethiopia's $3.4 billion Extended Credit Facility was completed in January; the fifth review reached a staff-level agreement in June and was approved in July, clearing another $468 million for disbursement. But the programme's credibility is undermined as long as a comprehensive restructuring remains unfinished.

Ethiopia's debt stock is itself a puzzle. Public-sector external debt stood at approximately $31 billion in September 2024 and has risen to above $34 billion by early 2026, according to the Ministry of Finance. More than half is owed to multilateral institutions which are traditionally treated as preferred creditors and receive little or no haircut. Non-Paris Club official creditors, mostly China, account for roughly a quarter. Private creditors hold 16%, including the defaulted Eurobond. The restructuring so far has addressed a fraction of the total; the multilateral share, being off-limits, is not part of the exercise. That too is a design feature, not a bug: official development finance is politically untouchable, even when it crowds out private capital and leaves countries perpetually leveraged.

The Eurobond deal is not a victory for debt sustainability. It is a narrow bridge over one small part of Ethiopia's liabilities, leaving the largest creditors — the multilaterals and China — still to settle their terms. Ethiopia's total public debt, domestic and external combined, exceeded $52 billion by December 2025. Even a successful restructuring will not reduce that burden enough to make the country comfortably investable. It will merely clear the default flag and open a window for the IMF programme to work.

The deeper lesson is institutional. The Common Framework was meant to be the successor to the Paris Club's old bilateral coordination. Instead, it has proven to be a slower, more opaque version of the same thing, with the added complication of private creditors who play by different rules. Its fatal flaw is not malice but design: a committee that cannot bind its members, a comparability principle that cannot be enforced, and a process that takes years to complete bilateral deals one by one. Countries with strong legal teams and patient bondholders — Argentina, Venezuela before its collapse — know this. Countries without them, like Ethiopia, are stuck in the middle.

The IMF's programme offers a counterweight. Ethiopia floated the birr in July 2024, moving from a managed peg to a market exchange rate, and has since liberalised parts of its financial system. External reserves, at 0.7 months of imports two years ago, have improved to an estimated 2.1 months by the current fiscal year. Growth, projected at 9% for 2026, is robust but fragile. Reforms are real; the debt overhang is more so. Without a comprehensive settlement, they will be perpetually one shock away from stalling.

What should follow is not another layer of process but a fundamental rethink. The G20 Common Framework needs either a collective action mechanism — a version of the cross-default clauses that govern bond contracts — or an honest admission that official and private creditors will always play different games and that "comparability" is a fiction. Until then, the system will continue to reward the wrong behaviour: taxpayers will subsidise private recoveries, litigious funds will extract better terms than development banks, and countries like Ethiopia will wait years for relief that is never quite comprehensive. The deal with the Eurobond holders is a step. It is not, however, the architecture the world needs.