For over two decades, newsletters and white papers have predicted the euro would dethrone the dollar, then the yuan would, then BRICS would launch a rival currency, then gold would reclaim its throne. None of it happened — not because the predictions were stupid, but because they mistook a declining headline percentage for a collapsing foundation. The dollar's share of global reserves has fallen from 72% to 57% since 2001. Its share of FX trading, trade invoicing, offshore credit, and crisis-time safe-haven demand has barely moved at all. This letter explains why — and what would actually have to change for the streak to break.
Not investment advice. Data sourced from Federal Reserve "International Role of the U.S. Dollar" 2025 Edition, IMF COFER data Q3 2025, Atlantic Council Dollar Dominance Monitor March 2026, BIS Triennial Central Bank Survey 2025, J.P. Morgan De-Dollarization Research, IISS "The Future of Dollar Dominance" January 2026. All figures current as of June 2026.
Before discussing the dollar, it is worth remembering what was said about the euro. In the years immediately following its 1999 launch, a substantial body of serious economic forecasting argued the euro would rival or surpass the dollar as the world's primary reserve currency within a generation — a large, stable, multi-trillion-dollar economic bloc, a credible central bank, and a population larger than the United States. The 2008 financial crisis, which began in the US, was widely treated as the moment the transition would accelerate. The 2010-2012 eurozone sovereign debt crisis briefly reversed that narrative, then commentators returned to it once the crisis passed. Two decades later, the euro's share of global FX reserves sits at approximately 20% — essentially unchanged from where it has hovered for most of the period, more than double the next-closest rival, but still less than a third of the dollar's share.
The euro did not fail because the European Union is weak. It failed to dethrone the dollar because it never solved the one structural problem that mattered most: the eurozone has a shared currency but no shared treasury. There is no single, deep, liquid pool of euro-denominated sovereign debt that functions the way US Treasuries function — instead there are German bunds, French OATs, Italian BTPs, each carrying different credit risk, trading at different yields, issued by governments that do not stand behind each other's debt. A reserve currency requires a reserve asset behind it — something foreign central banks can buy in essentially unlimited size without moving the price. The eurozone, by design, cannot offer that. This is not a temporary gap that more political will closes. It is a structural feature of how the currency union was built, and it has held for twenty-five years.
The "dollar is declining" narrative rests almost entirely on one statistic: the dollar's share of global foreign exchange reserves has fallen from roughly 72% in 2001 to approximately 57% in late 2025. That decline is real and worth taking seriously. But reserves are only one of at least five distinct channels through which a currency exercises global dominance — and the other four have barely moved. Conflating "reserves share declined" with "the dollar is losing its dominant role" is the single most common analytical error in two decades of de-dollarisation commentary, and it is the error this letter exists to correct.
*Trade invoicing estimates vary by source and methodology; even the lower estimates leave the dollar far ahead of any rival.
Look at the pattern: the one metric that declined meaningfully is the one that requires the least ongoing economic activity to change. A central bank can shift its reserve composition with a few large transactions over a period of months — it is a portfolio allocation decision. Trade invoicing, FX trading turnover, and debt issuance require millions of individual companies, banks, and traders to simultaneously decide that pricing and settling in a different currency is worth the cost and risk of switching — and that has not happened at anywhere near the same pace, for reasons explored below.
The US Treasury market is the deepest, most liquid sovereign debt market on earth — a central bank or sovereign wealth fund can buy or sell tens of billions of dollars of Treasuries in a single session without materially moving the price. No other government bond market offers this at comparable scale. China's onshore bond market remains subject to capital controls that prevent the free in-and-out flow foreign reserve managers require. The eurozone's bond market is fragmented across national issuers with different credit profiles, as explained above. Japan's JGB market, the subject of Letter 98's analysis, is itself under stress from its own debt dynamics.
A reserve manager holding dollars is not making a statement of confidence in the United States as a country. They are making a decision about where they can deploy and withdraw very large sums of money instantly, at any time, including during a crisis — and on that narrow, practical question, the Treasury market remains in a category of its own.
Because so much global trade is invoiced in dollars, companies hold dollar balances to pay suppliers and receive payment. Because companies hold dollar balances, banks build dollar-denominated services, credit lines, and trade finance products around them. Because banks offer those services, more trade gets invoiced in dollars, because it is simply the path of least friction. This is the "dominant currency paradigm" the IMF's Gita Gopinath has documented — the dollar shapes pricing even in trade between two countries that have no direct economic relationship with the United States, because both sides find it easier to price and settle in the currency everyone else already uses.
Switching away requires not one decision but millions of simultaneous, coordinated decisions across companies, banks, and counterparties who all have to trust that everyone else is switching too — otherwise an individual company that switches first simply bears extra cost and risk for no benefit. This is precisely the kind of coordination problem that resolves very slowly even when every individual participant might prefer the end state.
The single most counterintuitive fact about the dollar's dominance is this: in moments of acute global financial stress — 2008, March 2020 — demand for dollars spikes upward, even when the crisis originates inside the United States itself. The Federal Reserve is the only central bank in the world able to provide dollar liquidity at the scale the entire global financial system needs during a crisis, and it did exactly that in 2008 and 2020 through swap lines with other central banks, because so much of the world's trade finance, bank lending, and corporate debt is dollar-denominated and needs to be serviced in dollars regardless of where the crisis started.
A genuine reserve currency rival needs to demonstrate it can perform this function too — provide liquidity to the rest of the world during a global panic, not just accumulate reserves during calm periods. No alternative currency, including the yuan, has been tested in this role at scale, and China's capital controls make it structurally difficult to provide it even if Beijing wanted to.
None of this means nothing is happening. Real, measurable diversification is underway, and an honest letter does not pretend otherwise. The Shanghai Cooperation Organisation reported that 97% of trade among its ten member nations in 2025 was settled in local currencies rather than dollars or euros — a genuine and significant shift for that specific trade corridor. China's renminbi share of China's own goods trade settlement rose from roughly 18% in early 2023 to around 30% by Q3 2025. Russia-China bilateral trade is now over 90% settled in renminbi or roubles — though this is substantially a function of Western sanctions cutting Russia off from dollar and SWIFT access, not a voluntary preference that would hold absent sanctions. mBridge, the cross-border CBDC platform covered in Letter 100, had processed a cumulative $55.49 billion in transactions by November 2025 — meaningful, but a rounding error against the $1.9 trillion CHIPS alone settles every single business day.
Central banks have also been buying gold at the fastest pace in decades — 863 tonnes in 2025 alone, pushing gold's share of reserves above 20% in developed markets for the first time since the 1970s. This is real diversification behaviour, but gold is not a substitute for a reserve currency in any functional sense — it cannot be used to settle trade invoices, cannot back a bond market, and pays no yield. It is a hedge, not a replacement. Foreign ownership of US Treasuries has also genuinely declined, from over 50% during the 2008 crisis to around 30% by early 2025 — though this reflects the Treasury market's own rapid growth at least as much as active foreign selling, and Japan alone still holds over $1.1 trillion.
For a genuine, structural shift away from dollar dominance — not marginal diversification at the edges, but a true regime change comparable to sterling's decline after World War II — at least three conditions would need to be met, and none currently is. First, a credible rival needs a deep, liquid, freely convertible bond market of comparable scale to US Treasuries — the eurozone would need genuine fiscal union and common debt issuance at scale, something Letter 98's analysis of European bond market fragmentation already shows is not close; China would need to abandon capital controls entirely, which conflicts directly with the Chinese Communist Party's preference for controlling capital flows as a tool of domestic economic management.
Second, that rival needs demonstrated crisis-liquidity capability — the ability and willingness to flood the global financial system with its currency during a genuine panic, the way the Fed did in 2008 and 2020. No alternative has been tested at this scale, and the structural features that make the euro and yuan attractive in normal times (the ECB's mandate constraints, China's capital controls) work directly against this capability in a crisis. Third, and least discussed: network effects need decades to unwind even once the first two conditions are met, because millions of existing contracts, credit lines, invoicing relationships, and institutional habits do not reprice overnight — they amortise out gradually as new contracts get written, which is precisely why every "dollar killer" prediction for two decades has consistently overestimated the speed of the shift even when it correctly identified its direction.
US fiscal and political dysfunction is the most credible threat to dollar dominance — not foreign competition. The dollar's advantages are structural and durable against external rivals, but they are not immune to self-inflicted damage. Threats to Federal Reserve independence, erratic tariff policy that triggered a measurable dollar selloff in April 2025 against textbook predictions, public musings about engineering deliberate dollar depreciation, and a deteriorating long-term fiscal trajectory (directly connected to Letter 99's sovereign debt refinancing analysis) are all genuine vulnerabilities that originate in Washington, not Beijing or Brussels. The dollar has survived every external challenger for two decades. Its largest risk going forward is whether the United States manages its own institutions and finances in a way that preserves the trust the system depends on.
This is a gradual, multi-decade trend, not a stable equilibrium — and the rate of change matters more than the direction. Every individual metric explored in this letter has moved at the margins over the past two decades: reserves down meaningfully, everything else down slightly or flat. A rational, non-emotional forecast says this gradual erosion likely continues for the next two decades as well, without producing the dollar's "collapse" or "replacement" that newsletters keep predicting. The honest forecast is neither "nothing changes" nor "the dollar is dying" — it is "the dollar's dominance erodes slowly at the margins for decades while remaining structurally unrivalled at the core," which is a far less dramatic and far more accurate story than either extreme.
A multipolar currency world is being built in parallel, not as a replacement. mBridge, CIPS, SPFS, the BRICS Cross-Border Payments Initiative, and bilateral local-currency trade arrangements are not failed attempts at dollar replacement — they are successful builds of alternative infrastructure for specific corridors (sanctioned economies, intra-Asian trade, commodity discounting) that genuinely reduce dollar dependency in those narrow channels without threatening the dollar's role in the broader global system. This is worth tracking carefully and is genuinely consequential for the specific countries and trade flows involved — it is simply not the same story as "the dollar's global dominance is ending."
Long-horizon thinking on capital, technology, and the forces shaping the next decade of wealth creation. Written from first principles. Not consensus. Not noise.