Macroeconomics Without Empire: Designing a Pluralist Global Order

Imperial monetary orders share a structural logic. A single centre accumulates credibility over time, develops deep and liquid markets for its liabilities, and acquires the institutional capacity to act as a backstop when the system encounters stress. The sterling order, centred on London until its gradual dissolution across two world wars, provided that function for a century of industrial capitalism. The dollar order, formalised at Bretton Woods in 1944 and reformulated after 1971 on the basis of dollar credit rather than gold convertibility, has provided it since. The asymmetry is inherent to the structure. Peripheral economies accept the costs of dollar dependence — exposure to Federal Reserve monetary cycles, vulnerability to dollar liquidity crises, accumulation of reserves in assets whose security ultimately depends on political decisions in Washington — in exchange for access to a payment and reserve system of unmatched depth and liquidity.

That exchange is not collapsing. The dollar remains the currency of approximately half of all global trade invoicing, 47.48 per cent of SWIFT payment messages by value and approximately 47 per cent of global cross-border bank lending according to BIS, SWIFT and IMF data for 2023–24. The US Treasury market at $27.9 trillion in marketable debt outstanding is the deepest sovereign bond market in the world and the one in which stress relief, flight-to-quality flows and reserve management converge during crises. None of the alternative arrangements described in this dossier approaches those numbers. What has changed is not the dollar’s dominance but the perception of its permanence and, more precisely, of the conditions under which it can be relied upon.

The IMF’s COFER historical series shows the dollar’s share of global allocated foreign exchange reserves declining from 71.13 per cent in 2000 to 65.73 per cent in 2015 to 58.22 per cent in Q3 2024. The average annual rate of erosion is approximately 0.54 percentage points per year. At that pace, the dollar remains the dominant reserve currency through the 2030s and into the 2040s without any structural break in the current system. The decline is real; it is also slow.

The composition of what has replaced lost dollar share is instructive. The euro, at 19.69 per cent, is the primary beneficiary of reserve diversification, reflecting the depth of eurozone sovereign markets and the currency’s institutional credibility despite periodic stress. The renminbi, despite 15 years of deliberate internationalisation effort by China, holds 2.14 per cent. The remainder is distributed across yen, sterling, Canadian and Australian dollars and minor currencies. What this distribution shows is that reserve managers diversifying away from the dollar are not diversifying toward China. They are diversifying toward other established reserve currencies and, as the gold data shows, toward the one asset that carries no counterparty risk at all.

The dollar’s role in trade invoicing and payment systems is even more persistent than its reserve share. Approximately 50 per cent of global trade is invoiced in dollars according to BIS and ECB research, a figure that substantially exceeds the United States’ share of world trade, reflecting the dollar’s function as a vehicle currency in transactions involving neither an American buyer nor seller. SWIFT data for August 2024 shows the dollar accounting for 47.48 per cent of payment messages by value. These payment and invoicing functions generate the dollar demand that sustains reserve accumulation and create the network externality that makes switching costly even when geopolitical incentives to do so exist.

Central bank net gold purchases reached 1,082 tonnes in 2022, 1,037 tonnes in 2023 and are estimated at 1,010 to 1,040 tonnes in 2024, according to the World Gold Council’s Gold Demand Trends series. The historical context is significant. The previous peak in central bank gold accumulation was approximately 1,400 tonnes in 1967, during the collapse of the London Gold Pool, the last serious institutional attempt to maintain dollar-gold convertibility. Three consecutive years above 1,000 tonnes of central bank gold purchasing represent the highest sustained acquisition pace in the post-Bretton Woods era. The message is not subtle.

Gold does not pay interest. It does not mature. It cannot be frozen by a foreign government. It cannot be subject to secondary sanctions that restrict its use as a settlement asset. It does not require a SWIFT message to transfer title. In an era when the weaponisation of payment infrastructure and reserve asset custody has become a documented feature of geopolitical competition, gold’s attributes as a non-counterparty reserve asset have acquired a premium that pure yield calculations did not previously justify. The top buyers across 2022 to 2024 are the World Gold Council’s data: China at approximately 400 tonnes cumulative, Turkey at approximately 150 tonnes and India at approximately 75 tonnes. All three are economies that have reason to be attentive to custodial risk and sanctions exposure. None is signalling imminent departure from the dollar system. All are building sovereign balance sheet positions that do not depend on decisions made in Western financial capitals.

The Federal Reserve’s dollar swap line network is the closest thing the current monetary order has to a global lender of last resort mechanism for dollar liquidity. When dollar funding markets seized in March 2020 as the pandemic triggered a global flight to dollar liquidity, the Federal Reserve activated swap lines with 14 central banks, providing peak outstanding support of $448.7 billion according to the Fed’s H.4.1 statistical releases. That intervention stabilised global dollar markets within weeks. It was effective. It was also not universally available.

The Federal Reserve maintains permanent standing swap lines with five central banks: the Bank of Canada, the Bank of England, the Bank of Japan, the European Central Bank and the Swiss National Bank. All five are central banks of advanced economies, all five are G10 members, and none is an emerging market. The 14 counterparties in the 2020 temporary arrangements included a broader set, including Australia, Denmark, South Korea and Sweden, but remained confined to economies with established relationships and perceived systemic importance. No Sub-Saharan African central bank, no South Asian central bank and no significant Global South economy outside South Korea had access. In the most acute dollar liquidity crisis since 2008, the architecture of emergency support reproduced the hierarchy of the order it was designed to stabilise.

On 26 February 2022, G7 governments announced the immobilisation of approximately $300 billion in foreign exchange reserves held by the Central Bank of Russia, in response to the full-scale invasion of Ukraine. The action was unprecedented in scale. Central bank reserves had been subject to sanctions before, in smaller amounts, but the immobilisation of a major economy’s entire externally held reserve position in a single weekend fundamentally altered the perception of sovereign asset security for reserve managers worldwide. The largest single holder was Euroclear in Belgium, holding approximately €191 billion of the immobilised assets according to Euroclear’s own annual results and EU Commission reports for 2024.

The economic consequences of the immobilisation have accumulated since then. Euroclear’s financial statements for 2024 record €5.1 billion, approximately $5.5 billion, in net profit generated from the frozen assets through interest earnings on reinvested cash collateral. In June 2024, the G7 Apulia Summit agreed an Extraordinary Revenue Acceleration loan of $50 billion to Ukraine, to be serviced from the stream of interest income generated by the frozen Russian assets, which will continue to accrue for as long as the assets remain immobilised. The loan and its servicing mechanism represented a further step in the institutionalisation of reserve asset weaponisation: not merely freezing assets but actively deploying their financial returns as war financing.

For central banks worldwide, the lesson was not primarily about Russia. It was about the conditions under which reserve assets held in the infrastructure of the Western financial system can be treated as subject to political decision. China, which was simultaneously reducing its US Treasury holdings from approximately $1.05 trillion to approximately $770 billion between 2022 and 2024 according to US Treasury International Capital data, accelerating its gold purchases to approximately 400 tonnes cumulative over the same period and expanding CIPS transaction volumes, read the Russian experience as a demonstration of its own potential exposure. The strategic portfolio shift did not require any announcement. It was visible in the data.

China’s Cross-Border Interbank Payment System processed 123.1 trillion Chinese yuan, approximately $17.3 trillion, in transaction volume in 2023–24 according to PBOC and CIPS Annual Operating Report data, with over 1,500 participating financial institutions across more than 100 countries. These are substantial numbers. The honest comparison is with SWIFT, which processes approximately $150 trillion in annual transaction volume, representing a daily volume of roughly $42 trillion per quarter. CIPS handles approximately 11.5 per cent of SWIFT’s annual volume and serves approximately 1,500 institutions compared with SWIFT’s 11,000 plus. CIPS is growing. It is not a systemic alternative to SWIFT. It is a parallel channel that provides redundancy for CNY-denominated transactions and reduces Chinese financial system dependence on SWIFT infrastructure without displacing it.

Project mBridge, developed through the BIS Innovation Hub with the central banks of China, Thailand, the United Arab Emirates, Hong Kong and, more recently, Saudi Arabia, reached Minimum Viable Product status in 2024 according to BIS Innovation Hub updates. The project tests multi-currency central bank digital currency settlement on a shared platform, enabling direct central bank-to-central bank transactions that bypass commercial correspondent banking and the SWIFT messaging network. At MVP stage, mBridge remains a pilot rather than an operational system. Its significance is directional rather than current: it demonstrates that technically viable alternatives to SWIFT-based settlement exist and are being developed by central banks that represent a significant share of global trade flows.

India’s rupee trade settlement initiative had opened Special Rupee Vostro Accounts with Indian banks for 22 countries by 2024 according to RBI Annual Report data. The mechanism allows bilateral trade between India and partner countries to be settled in rupees through accounts held at Indian banks, reducing dollar denomination requirements in specific trade corridors. The volume is growing but remains modest relative to India’s total trade. Renminbi’s share of global trade finance through SWIFT stood at 5.28 per cent in August 2024 according to the SWIFT RMB Tracker, reflecting the currency’s growing but still limited role in global trade documentation and financing.

The renminbi’s share of global allocated foreign exchange reserves stands at 2.14 per cent according to IMF COFER data for Q3 2024. This number is both a measure of progress and a measure of constraint. Progress: the renminbi was not included in COFER reporting as a separate category until 2017, reflecting its admission to the IMF Special Drawing Rights basket in 2016, and its share has grown from a minimal base. Constraint: despite a $17.3 trillion annual CIPS payment volume, the mBridge project at Minimum Viable Product stage, a 22-country rupee settlement network in parallel with China’s own bilateral currency arrangements, and a BRICS group that represents 36.7 per cent of world GDP by purchasing power parity, the renminbi holds 2.14 per cent of global reserves against the euro’s 19.69 per cent and the dollar’s 58.22 per cent.

The explanation is not political will. It is capital market architecture. Currency internationalisation requires three structural conditions that are mutually reinforcing: a deep and liquid domestic bond market that can absorb large non-resident positions, credible and predictable monetary policy that makes the currency reliable as a store of value, and open capital account arrangements that allow non-residents to move in and out of renminbi assets without restriction or unpredictable intervention. China meets the first condition substantially. Its monetary policy credibility is established but not fully independent in the sense that international reserve managers require. Its capital account remains significantly controlled. Reserve managers diversifying toward renminbi face the practical question of what they would do with large positions in a stressed scenario where capital account restrictions might prevent liquidation. Until that question is answered by the architecture rather than by assurances, the renminbi’s reserve share will reflect the capital market reality rather than the geopolitical ambition.

Emerging market portfolio flows experienced net outflows of $14.1 billion in 2022 according to Institute of International Finance Capital Flows Tracker data, as the Federal Reserve’s tightening cycle transmitted to emerging markets through yield differentials, dollar appreciation and risk repricing. The JPMorgan and BIS EM Currency Index declined 8.5 per cent against the dollar during 2022. The transmission mechanism of advanced-economy monetary policy to emerging market financial conditions is one of the most persistent structural features of the current monetary order, and one of the most politically contested.

The structural response has been the deliberate development of local currency bond markets. Emerging market local currency bonds outstanding reached $35.2 trillion according to BIS and World Bank Quarterly Review data for 2024, representing two decades of institutional deepening in sovereign debt management, pension fund regulation and yield curve development. The figure is large and it represents genuine progress. It has not solved the core vulnerability: 67 per cent of emerging market external debt remains denominated in foreign currency according to the World Bank International Debt Report 2024. The large local currency market exists alongside a large foreign currency exposure. The two figures together describe a structural duality in which domestic capital market deepening has not yet displaced the foreign currency borrowing that remains necessary to finance current account deficits and infrastructure investment at competitive cost.

The G20 Independent Expert Group, co-chaired by Mia Mottley and Larry Summers, published its Triple Challenge report in 2023 recommending that multilateral development banks expand their annual lending capacity by $260 billion to address the combined demands of development finance, climate transition and pandemic preparedness. The World Bank’s own Evolution Roadmap, published in 2024, estimated that capital adequacy reforms would generate $40 to $50 billion in additional IBRD lending headroom over ten years. The gap between the G20 IEG recommendation and the World Bank’s own capacity estimate is approximately $210 billion per year. That gap is the measure of the institutional reform that has not yet occurred.

The blended finance mobilisation ratio published by the OECD in its 2023 Blended Finance Report stands at $0.37 in private capital mobilised per dollar of MDB or public capital deployed. Less than 40 cents on the dollar. The premise of blended finance architecture is that public capital can crowd in private capital at multiples of the original investment, reducing the demand on public budgets and MDB balance sheets. The $0.37 figure indicates that the system is not working at the ratios that its proponents anticipated. The IMF’s 16th General Review of Quotas, agreed in 2023, increased total quotas by 50 per cent to 715.7 billion SDRs, strengthening the Fund’s resource base. The global financial safety net, encompassing foreign exchange reserves, central bank swap lines, regional arrangements and IMF resources, has been estimated at approximately $15.8 trillion by the IMF and Financial Stability Board. It is large in absolute terms. It is asymmetrically distributed and access-constrained in ways that its aggregate size does not reveal.

The BRICS grouping admitted four new members in January 2024 — Egypt, Ethiopia, Iran and the United Arab Emirates — with Saudi Arabia participating without full formal membership, following the Johannesburg II and Kazan declarations. The expanded BRICS+ collectively represents approximately 36.7 per cent of world GDP at purchasing power parity and approximately 25 per cent of world trade according to IMF World Economic Outlook October 2024 and WTO/IMF trade data. Thirty-four countries had expressed formal interest in BRICS membership by 2024 according to the BRICS Secretariat and South African Foreign Ministry. These are substantial numbers. They reflect a genuine reorientation of political economy preferences in the Global South toward institutions that are not centred on Western governance frameworks.

The financial architecture of BRICS tells a different story. The BRICS Contingent Reserve Arrangement, the grouping’s principal monetary safety net instrument, has a total committed size of $100 billion. The Chiang Mai Initiative Multilateralisation, the ASEAN+3 equivalent, is $240 billion. The IMF’s total lending capacity is approximately $1.3 trillion. The BRICS CRA has been called upon rarely and its effectiveness in a major crisis has not been tested. Moreover, 70 per cent of CRA access above a threshold requires IMF programme conditionality, meaning that the BRICS financial safety net is substantially a referral to the institution whose governance the grouping most vocally critiques. The political weight of BRICS+ and the financial weight of its institutional architecture are not yet in correspondence. That gap, between 36.7 per cent of world GDP and $100 billion in reserve arrangement capacity, is the most precise measurement available of the distance between ambition and architecture in the pluralist project.

A pluralist monetary order is not an order without architecture. It is an order whose architecture is distributed across multiple institutions, currencies and governance frameworks rather than concentrated in a single hegemonic centre. The risk of unmanaged multipolarity is not merely instability but incoherence: payment systems that cannot interoperate, capital adequacy standards that diverge across blocs, and debt resolution mechanisms that cannot achieve coordinated outcomes because no single creditor has the leverage to convene the others. The fragmentation of sovereign debt restructuring processes under the G20 Common Framework — where Zambia’s restructuring required years of negotiation between Chinese bilateral creditors, Western commercial bondholders and multilateral institutions, each operating under different mandates and incentives — illustrates what incoherence costs in practice.

The requirements for a stable pluralist order are therefore institutional rather than hegemonic. Payment system interoperability must be maintained even as alternative networks develop. mBridge at MVP stage and CIPS at 11.5 per cent of SWIFT volume are not threats to systemic stability. They become potential sources of fragmentation if their governance frameworks diverge from global standards in ways that prevent cross-network settlement. Capital adequacy frameworks that apply only within blocs create regulatory arbitrage. Reserve assets diversified toward gold and non-dollar alternatives reduce counterparty risk while maintaining the liquidity requirements that make reserves usable in a crisis, requiring reserve managers to think carefully about the composition rather than simply the direction of diversification.

The world is not post-empire. The dollar accounts for 58.22 per cent of global allocated reserves, 47.48 per cent of SWIFT payment values and approximately half of all global trade invoicing. The US Treasury market at $27.9 trillion remains the singular depth benchmark for sovereign fixed income. In the next systemic dollar liquidity crisis, the Federal Reserve will again be the only institution capable of extending swap support at sufficient scale to stabilise global funding markets. None of the alternative architectures — CIPS at 11.5 per cent of SWIFT volume, BRICS CRA at $100 billion, mBridge at MVP stage, the renminbi at 2.14 per cent of reserves — changes that. What has changed is the perceived reliability of the conditions under which the dollar system extends its benefits to all participants equally. The $300 billion Russian reserve freeze was the most consequential single event in reserve management since the end of Bretton Woods, not because it changed the arithmetic of dollar dominance but because it demonstrated that dollar dominance includes the power to exclude.

The rational response of Global South economies to that demonstration is not to exit the dollar system. It is to reduce marginal exposure to the most politically sensitive elements of it: custodial concentration in Western financial infrastructure, dollar-denominated external debt where alternatives exist, dependence on SWIFT for bilateral transactions in local currencies where CIPS or direct settlement is feasible. These adjustments are underway. They are incremental, rational and visible in the data. Central bank gold at 36,150 tonnes globally, three years of 1,000-plus-tonne purchases, India’s $701 billion in reserves with 854 tonnes of gold, 22 rupee settlement corridors, mBridge at MVP: the direction is consistent even as the scale remains modest relative to the system being hedged against. The transition is real. Its tempo is evolutionary, not revolutionary.