Plain-language explainers on the forces shaping the world.
Reserve currency status survives long after the issuer's economic share has fallen. The reason is liquidity and habit, not strength.
Central banks hold dollars because other central banks hold dollars. Coordination problems do not unwind quickly.
USD share of global FX reserves: ~59% (IMF COFER, Q4 2023)
Mark Sobel, the former US Treasury deputy assistant secretary and now US chair of OMFIF, plus Brookings's Homi Kharas, have produced the most policy-grounded analysis of the IMF's August 2021 special-drawing-rights allocation. The headline $650B number conceals a sharply skewed distribution.
An SDR allocation is a non-disbursable reserve asset. For most rich countries that means a balance-sheet line item. For most poor countries it means a real if modest budget-financing boost. The political fight is about transfer between the two.
2021 SDR allocation: SDR 456B ($650B at then-exchange rates). Share allocated to G7 economies: about 63%. Share to low-income countries: about 3% (IMF, OMFIF analysis).
Stefan Avdjiev and Patrick McGuire at the Bank for International Settlements have spent two decades mapping the offshore dollar system — bank balance sheets denominated in US dollars but booked outside US borders. The headline finding is that this system dwarfs the Federal Reserve's own balance sheet.
The dollar is not a currency the Fed prints. It is a unit of account that foreign banks create through credit, and the Fed is the only backstop when the credit unwinds.
Non-US banks' US-dollar liabilities: about $13 trillion in 2024 (BIS Quarterly Review, Avdjiev and McGuire tracking).
The US dollar is used in roughly 88% of foreign exchange transactions and over half of all trade invoicing. That dominance isn't an accident — it's a network effect that compounds even when US economic share shrinks.
Every time a Brazilian importer pays a Chinese supplier in dollars, the dollar's network gets a little stickier.
USD share of FX transactions: 88% (BIS, 2022 Triennial Survey)
David Spiro's 1999 book *The Hidden Hand of American Hegemony* remains the canonical account of the 1974 Saudi-US arrangement that channeled OPEC oil revenue into US Treasuries. Fifty years on, the arrangement is being quietly amended.
The petrodollar was never a single signed deal. It was a set of overlapping arrangements that lasted because the alternative was worse for both sides. The alternative is now less worse.
Share of Saudi crude sales settled in non-USD currencies: estimated 1-2% in 2022, 5-7% in 2024 (industry estimates, Reuters and Energy Intelligence reporting).
Adam Tooze at Columbia and Daniela Gabor at UWE Bristol have argued that covered interest parity breakdowns since 2008 are the most important under-discussed fact in international finance. The Federal Reserve's swap lines are the institutional answer to a question the textbook does not ask.
Covered interest parity is the law that ties the dollar system together. When it breaks, the Fed has to fix it. When the Fed has to fix it, the Fed is the world central bank.
Peak Fed swap-line drawdowns: $583B in December 2008; $446B in May 2020 (Federal Reserve H.4.1 release; Tooze, Crashed, 2018).
Local-currency settlement systems are multiplying — PAPSS in Africa, mBridge in Asia, INSTEX (now defunct) in Europe. The marketing talks about ending dollar dominance. The plumbing changes are slow, partial, and structurally bounded by what local currencies can credibly do.
A payment rail is only as useful as the currency liquidity behind it. Bridges between thin currency markets remain thin currency markets.
PAPSS settled $1.2B in cross-border African trade in 2024; SWIFT settles ~$10T daily globally.
Eswar Prasad at Cornell, in *Gaining Currency: The Rise of the Renminbi* (Oxford, 2017) and his follow-up work, has tracked the gap between China's growing trade-invoicing share and its much smaller share of global reserves. The gap is the story.
China can settle trade in renminbi. China cannot make foreign central banks hold renminbi as reserves. The first is a policy choice. The second requires open capital markets China will not deliver.
RMB share of global trade invoicing: about 7% in 2024. RMB share of global FX reserves: about 2.4% (IMF COFER 2024Q3; SWIFT RMB Tracker).
Carla Norrlof at the University of Toronto, in *The Global Power of the US Dollar* (Cambridge, 2024) and her earlier work, has made the cleanest analytical distinction between financial messaging systems and settlement systems. The popular framing of SWIFT as a US weapon misses where the real power sits.
SWIFT is a phone network. The settlement happens in dollars, in New York, on Fedwire and CHIPS. Cutting the phone does not move the money. Cutting access to the money does.
SWIFT messages per day: about 50 million. CIPS direct and indirect participants: about 1,500 banks across 110 jurisdictions (SWIFT, PBoC data, 2024).
Hyun Song Shin, economic adviser and head of research at the Bank for International Settlements, has argued that dollar stablecoins are best understood as a private offshore-dollar system — a digital eurodollar with a money-market-fund balance sheet and no backstop.
If it pegs to a dollar, holds short-dated dollar assets, and promises redemption on demand, it is a money-market fund. The question is which money-market fund regulations apply when nobody calls it a money-market fund.
Stablecoin market capitalization (2024): about $160B, of which Tether (USDT) $115B and Circle (USDC) $35B (CoinGecko, BIS tracking).
The IMF's 2023 working paper on central bank digital currency interoperability, combined with the Atlantic Council's CBDC Tracker, gives the clearest map of where CBDC projects actually sit. The hype gap between announcement and operational issuance is the policy story.
A CBDC is not a threat to the dollar. A multi-CBDC settlement rail — that bypasses the US correspondent system — might be. The two are constantly confused.
Countries with launched retail CBDCs (mid-2024): 4 (Bahamas, Jamaica, Nigeria, Eastern Caribbean). Countries in pilot: 36. Countries in research only: 94 (Atlantic Council CBDC Tracker).