The Dollar and the Monetary Order
The slow erosion of reserve-currency privilege, the rise of parallel payment systems, and what happens when sanctions become a routine tool of foreign policy.
(down from about 71% in 2000)
(about $620 billion of it by China)
(little changed since the 2000s)
A note on framing. "The dollar is collapsing" and "the dollar will dominate forever" are both wrong, and both are popular. The data tells a calmer story: the dollar's privileged role is real, the erosion of that privilege is also real, and most of the change is happening on a slow geological timescale rather than a breaking-news one. The page below tries to show the actual shape of that change.
What "reserve currency" means
A reserve currency is the money other countries' central banks choose to hold as savings. When the People's Bank of China or the European Central Bank or the Central Bank of Brazil sets aside foreign-currency savings to back its own currency, defend against shocks, and pay for imports during emergencies, those savings are mostly held in US dollars. About 57% of all such reserves worldwide were held in dollars in early 2026. The rest is mostly euros, with smaller shares for Japanese yen, British pounds, Chinese yuan, Canadian dollars, Australian dollars, and Swiss francs.
Why does any of this matter to a non-economist? Because the dollar's status as the world's main reserve currency creates real, measurable advantages for the United States and real, measurable constraints on everyone else. American companies can buy and sell goods abroad in their own currency. American consumers borrow at lower interest rates than they otherwise would because the world has a permanent appetite for dollar-denominated savings. American sanctions can shut a target country out of the dollar payment system - which means out of most of the world economy. The French finance minister in the 1960s called it "the exorbitant privilege" of the dollar, and the phrase has stuck because nobody has come up with a better description.
For a long time after 1945, this arrangement was largely unchallenged. The dollar replaced sterling as the central reserve currency in the early 20th century, and the Bretton Woods agreement of 1944 cemented the role by pegging other currencies to the dollar and the dollar to gold. Even after the United States abandoned gold convertibility in 1971, the dollar's role as the world's saving and trading currency stayed essentially unchanged. That world is the one most readers grew up in. It is now changing - slowly, partially, and not for any single reason.
Who is still buying US debt
The single cleanest signal of dollar trust is whether the rest of the world keeps buying US Treasury debt. Foreigners held about $9.2 trillion of it in July 2026 - safe, liquid government securities that any major central bank can buy or sell almost instantly. That number has been growing for fifty years, with one important change in the trend.
Foreign holdings rose steadily from the early 1970s through about 2014, then flattened for four years. Since 2019 the dollar value has grown again, but as a share of US debt, foreign ownership has kept declining. At the end of 2014, foreigners owned nearly half of US debt held by the public; by July 2026 they owned about 29%. The gap has been filled by domestic American institutions and the Federal Reserve. The world is not selling US debt - it is just not buying it as fast as the United States is issuing it.
Inside that headline number, the country mix has also shifted. China was the largest foreign holder for about a decade and has been steadily reducing its position since 2013, from roughly $1.3 trillion to about $620 billion. Japan, at about $1.1 trillion, is again the largest foreign holder. The United Kingdom - most of which is really international money parked in London - moved ahead of China in 2025, leaving China third. Saudi Arabia and the United Arab Emirates have added to their holdings; India and Brazil hold less than in early 2023. Since then almost all the growth has come from private investors; foreign governments now hold about 41% of the foreign total, down from half. The buyer base has become more diverse, less Asian-government-driven, and more market-driven - which probably makes it more responsive to short-term price changes than it used to be.
How the dollar has moved
People who talk about the dollar "collapsing" usually mean its value falling against other currencies. By that simple measure, the dollar has been historically strong rather than weak over the last decade.
The chart tracks how strong the dollar is against the basket of currencies the US actually trades with (it is not adjusted for inflation). The takeaway: after weakening in the late 2000s, the dollar ran up for a decade from 2014, surged again in 2022, and peaked in January 2025. It gave back about 8% by mid-2025 and has moved in a narrow band since, rising about 2% in the two weeks after US and Israeli strikes on Iran began in February 2026. A reader picturing a long dollar collapse should compare that with this chart. The recent turn is real and visible; it comes after a decade of the opposite.
Two reasons explain the strength. First, the United States kept growing faster than most other developed economies post-2008, which attracted savings from abroad. Second, when the world is uncertain (financial crisis, pandemic, war), money flows to dollars as the safest asset available. That "flight to safety" is exactly the reserve-currency advantage in action: at the moment when people need protection most, they buy dollars, which lifts the dollar's value, which makes it cheaper for Americans to import goods and harder for American exporters to compete. The privilege costs the export-heavy parts of the US economy and benefits the consumer-heavy parts. Both are true at the same time.
That long run broke in 2025. The most-quoted dollar index, a narrower gauge than the chart's, fell close to 11% in the first half of the year, its steepest first-half decline since 1973. Analysts cited tariff policy, public pressure on the Federal Reserve over interest rates, and doubts about the fiscal path. The same spring, Moody's cut the US long-term rating from Aaa to Aa1. It was the last of the three major agencies to remove the top grade; Standard & Poor's did so in 2011 and Fitch in 2023.
Both events are routinely oversold, in both directions. A falling exchange rate is not the same thing as a currency losing reserve status. The dollar fell hard in the 1970s and again in the 2000s without losing its place as the main reserve currency. A cheaper dollar is also what the export-heavy parts of the US economy have been asking for. The downgrade is more interesting for its reasoning than its rating. Moody's attached a stable outlook and cited the dollar's reserve-currency role among the reasons. The status this page is about is part of what stopped the downgrade going further. That is the privilege doing exactly what the privilege does - and a reminder that it is the thing with the most to lose.
The Federal Reserve's balance sheet
One reason US interest rates stayed low for so long despite enormous government borrowing is that the Federal Reserve bought US debt on a vast scale. From 2008 to 2022, the Fed grew its balance sheet from about $1 trillion to about $9 trillion, mostly by buying Treasury bonds and mortgage-backed securities to keep interest rates low after the financial crisis and the pandemic. This is the policy known as "quantitative easing" - the central bank creating new dollars to buy long-term debt.
From 2022 the Fed ran this back slowly, which is called "quantitative tightening," shrinking the balance sheet to about $6.5 trillion before stopping on December 1, 2025. Since mid-December it has been buying Treasury bills again to keep bank reserves ample; total assets were about $6.7 trillion in September 2026. Each direction has costs: a smaller balance sheet pushes interest rates up and raises the government's interest bill; a larger one, critics argue, erodes the dollar's purchasing power and hurts savers.
For the rest of the world, this matters because the Federal Reserve's rates and balance sheet set the cost and supply of the dollars that other countries borrow and trade in. When the Fed expands, dollars become abundant globally and emerging-market borrowing booms. When the Fed contracts, dollars become scarce and emerging markets get squeezed. The Fed's decisions about its own balance sheet effectively set the monetary weather for half the planet, and its mandate is - explicitly - to make those decisions for American conditions, not for the world. This asymmetry is one of the things countries trying to reduce dollar dependence are reacting to, even when they cannot say so directly.
The currency map of the world's savings
The cleanest way to see the dollar's privilege is to look at the entire reserve-currency picture. The numbers below are from the IMF's quarterly survey of the currencies central banks actually hold (the COFER data set), as of the first quarter of 2026.
The pattern is more "the dollar gradually shares the field" than "the dollar collapses." About 14 percentage points of share have moved out of the dollar over twenty-five years - real but slow. Most of that has gone into a basket of smaller currencies rather than into a single rival, while gold has grown alongside. There is currently no second currency capable of absorbing dollar-scale reserve flows, which is the deepest reason the privilege persists despite the slow erosion at the edges.
The sanctions question
The biggest change to the dollar's role over the last decade has not been in any economic statistic. It has been in the active use of dollar access as a foreign-policy weapon. The United States has the legal and technical ability to remove a country, a company, or a person from the dollar payment system at short notice. After Russia invaded Ukraine in 2022, roughly half of Russia's foreign-currency reserves - about $300 billion - were frozen in the Western countries that held them. The lesson many other central banks drew was that reserves held in Western currencies are not unconditional savings; they are conditional on staying in the good graces of Washington and its allies.
Most countries are nowhere near being sanctioned, and most countries do not particularly care because they have no plans to do anything that would risk it. But the central banks of countries that might one day fall on the wrong side of US foreign policy - China, Saudi Arabia, India, Brazil, South Africa, and a long list of others - have spent the last few years diversifying. More gold. More trade settled in local currencies, as between Russia and China. More bilateral currency-swap agreements that bypass the dollar. More development of alternative payment systems (China's Cross-Border Interbank Payment System, Russia's SPFS messaging network, India's Unified Payments Interface extending across borders). BRICS governments have studied linking such systems since 2024, but their September 2026 declaration only "acknowledged" that work and proposed no common currency. None of these alternatives are large enough to replace the dollar system today. They do not have to be. They just have to provide a fallback option in the event of a future confrontation, and provide it for enough countries that the network effects start to compound.
What this means in practice: the dollar's deep advantages (size, liquidity, legal predictability, deep bond markets, safe-haven flows) all remain. It still carries about half of the payments sent over the SWIFT bank-messaging network, a share that has edged up in recent years. Its dominance is being whittled down only at the edges, mostly by countries that worry about being sanctioned. Whether the whittling stays at the edges or reaches the core depends on whether the United States continues to use sanctions as routinely as it has over the last twenty-five years.
The years since have added a second lesson that cuts the other way, and it is noticed far less often than the first. Freezing reserves proved easy. Spending them proved extremely hard. Of that frozen balance, more than half - on the order of €200 billion - sits at Euroclear, a securities depository in Belgium. The European Commission proposed lending Ukraine roughly €140 billion against that balance, structured as a reparations loan so that the money was borrowed against the assets rather than confiscated outright. It did not pass. Belgium, which hosts the depository and would absorb both the litigation and the retaliation, declined to carry a risk concentrated almost entirely on one small member state. Russia's central bank sued Euroclear in a Moscow court, which in May 2026 ordered Euroclear to pay it €200 billion; Euroclear contests the ruling. European leaders instead financed Ukraine with a €90 billion loan for 2026-27, raised on the capital markets against the EU's own budget, and made the freeze open-ended.
It is worth being precise about what decided that, because it was not either of the two arguments being made in public. Not the European position that an aggressor should pay for the damage it caused, and not the Russian position that this amounted to theft of sovereign property. What decided it was a third consideration neither side was arguing: the exposure sat in one jurisdiction, and that jurisdiction did not want it. Institutional plumbing beat both principles. For any central bank reading the episode, the signal is two-sided - reserves can be immobilized over a political rupture, quickly and with little difficulty, and converting them into someone else's money took years and has still not fully happened. Which of those two lessons a given government weighs more heavily is a reasonable thing for reasonable people to disagree about, and the diversification described above is what it looks like when they weigh the first one more.
The paths from here
Dated September 2026, with the dollar about 8% below its January 2025 peak, gold near $4,300 an ounce and the 10-year Treasury yield above 5%. None of the realistic paths involves the dollar's sudden collapse. Several involve its slow erosion, with very different distributions of who pays the cost. Each path includes a question about whether the political and structural conditions support it.
Slow erosion continues at current pace
The dollar's share of reserves keeps drifting down by about 1 point every two years, the basket of alternatives keeps growing, and parallel payment systems keep expanding without ever fully replacing the dollar. By the late 2030s the world is meaningfully more multi-currency than it is today, but the dollar is still first among many.
Will it happen? This is the base case and the path the data has been on since the early 2000s. The Iran war has not broken it so far; the dollar's share edged up in the first quarter of 2026. The path assumes no crisis centered on the United States itself, no sharp escalation of US sanctions, and no new credible reserve currency emerging. All three assumptions are reasonable but not certain.
A genuine multi-currency world
The dollar settles around 50% of reserves, the euro and yuan rise meaningfully, gold takes a permanent larger share, and a real basket of "non-aligned" currencies emerges for trade settlement among countries that do not want to choose between Washington and Beijing.
Will it happen? Possible by the 2040s. The two biggest barriers - the eurozone's lack of a unified safe-asset bond market, and China's capital controls - would need to be at least partly addressed. Neither change is impossible. Joint EU borrowing has grown since 2020, and international bond issuance in euros hit a record in 2025 (ECB), but neither barrier has come down. The realistic version of this path is "more currencies as supplements" rather than "the dollar loses its position."
Sanctions overreach accelerates the erosion
The United States imposes sanctions in another major case (a confrontation with China, a wider Middle East conflict, or further expansion of sanctions on Russia) in a way that pushes more countries to actively reduce their dollar exposure faster than the gradual baseline. The slow erosion turns into a faster one.
Will it happen? Possible, especially if there is a Taiwan crisis. A wider Middle East conflict has come with the Iran war, but Iran was already largely cut off from the dollar system, and reserve data through March 2026 show no faster shift. Each US administration since 2008 has used sanctions more aggressively than the last, and there is no obvious political incentive to slow that down. The deeper question is whether enough alternative infrastructure exists by then to absorb the displaced flows. Current alternatives are still small, but they are growing.
A US fiscal crisis shakes confidence
A serious US budget crisis - a debt-ceiling standoff that goes too far, an unexpected interest-rate shock, or an event that exposes the unsustainable trajectory of US debt - causes a one-time sharp drop in dollar trust. Reserves shift faster, the dollar weakens, and interest rates on US debt jump.
Will it happen? The 2022 UK government-bond crisis showed how fast confidence can shift even in a well-governed advanced economy. The US had a brief taste in April 2025, when the 10-year Treasury yield jumped from about 4.0% to 4.5% in a week after sweeping tariff announcements while the dollar fell - the reverse of the usual flight to safety. In September 2026, as the Fed raised rates against elevated inflation, the yield reached 5.18%. A full crisis in the US is unlikely but not negligible. The dollar's safe-haven status would partly cushion any move - paradoxically, in a crisis people often buy more dollars rather than fewer - but a crisis driven by the US itself rather than by external shocks is the version where the cushion may not work.
A new Bretton Woods moment
A major crisis (financial, geopolitical, or both) forces a coordinated reset of the international monetary system, with new rules about reserve currencies, sanctions, capital flows, and emergency liquidity. The current system gets restructured rather than slowly worn down.
Will it happen? Bretton Woods itself happened in 1944 because of a world war. A similar moment of consensus seems unlikely without a similar level of shock. Most analysts treat this as a low-probability tail rather than a base case, though Zoltan Pozsar and a small group of others have argued that something resembling it is already quietly underway.
The base case is slow erosion at roughly the current pace, with a steady rise of alternatives at the edges. There is a real but small probability of a sharper move, in either direction, depending mostly on US policy choices and on whether China makes the structural changes that would let the yuan absorb genuine reserve-scale flows. The next ten years are probably more boring than the loudest predictions on either side suggest.
Where serious analysts disagree
The dollar is one of the topics where careful disagreement is much more useful than confident prediction. Each of the readings below is held by named analysts with different vantage points, and the data does not cleanly settle any of them.
The dollar's role is far stickier than the headlines suggest
Network effects in money are stronger than in any other domain. Once the world's banks, businesses, and governments have built decades of contracts, systems, and habits around the dollar, switching is more like asking the world to change its common language than swapping a vendor. Even imperfect alternatives would have to overcome enormous coordination costs to displace it. The slow erosion of share is real, but the underlying privilege is much harder to dislodge than fans of dollar-decline narratives recognize.
Held by: Many central bankers and international economists. A 2025 review by Federal Reserve staff found the dollar in 58% of reserves, nearly nine in ten currency trades and about half of SWIFT payments, little changed since 2010, and judged it "will likely remain dominant" absent large-scale disruptions. In an ECB survey published in June 2026, over two-thirds of central banks weighing geopolitical risk had not changed their currency mix. The open question is whether sanctions overreach can break enough of those networks to overcome the coordination cost.
A new monetary order is quietly already here
Since 2022, the world has been running a parallel system in which gold, commodities, bilateral swap lines, and non-dollar payment rails are doing more of the work that dollar reserves used to do. This is not a future scenario. It is an ongoing transition that the official statistics under-count because they were designed for the old system.
Held by: Zoltan Pozsar (formerly Credit Suisse, now a private analyst) and a group of monetary historians who watch interbank flows rather than headline reserve numbers. The "Bretton Woods III" thesis has been controversial. Its supporters point to two facts: gold has overtaken US Treasuries as a share of official reserves at market prices, and 74% of central banks surveyed by the World Gold Council in 2026 expect the dollar's share of reserves to be lower in five years.
The yuan can't be a real reserve currency under capital controls
For the yuan to play the role the dollar plays, foreign central banks need to be confident they can move in and out of yuan-denominated assets at any size, on any timing, without the Chinese government changing the rules. As long as Beijing decides who can move money in and out of China, the yuan will stay below its trade-share weight in reserves. China would have to choose between domestic capital control and yuan internationalization, and there is no sign of that choice being made.
Held by: Eswar Prasad at Cornell and most working specialists on Chinese monetary policy. The data on yuan reserve share supports their position. Despite a generation of "yuan internationalization" announcements, the yuan's reserve share peaked just under 3% in 2021-22 and was about 2% in early 2026, far below China's roughly 14% share of world goods exports.
Sanctions are the real issue, and they are eroding the system from the inside
The dollar's exorbitant privilege has always rested on an implicit promise that holding dollars was politically neutral - a country could store its savings there and trust they would still be there when needed. The active use of sanctions as foreign policy has broken that promise. Each major sanctions episode since 2014 has accelerated the slow erosion, and sanctions are not slowing down. The privilege is not being lost to a competitor; it is being given away by Washington itself.
Held by: Daniel McDowell of Syracuse University, whose 2023 book Bucking the Buck links US financial sanctions to "anti-dollar" policies abroad. The evidence, by Russian accounts: almost all Russia-China trade is now settled in rubles and yuan, and about 96% of India's trade with Russia runs through rupee-ruble channels. McDowell himself doubts that sanctions alone will end the dollar's dominance.
Digital currencies will transform the field faster than anyone expects
Stablecoins, central-bank digital currencies, and tokenized real-world assets are reshaping international payments in a way most reserve-currency analysis does not yet account for. The unit-of-account question (what currency is X priced in) and the medium-of-exchange question (what currency is X paid in) are about to be partially separated by technology, with consequences nobody has fully thought through.
Held by: Christian Catalini (formerly with the Diem project) and a number of technologists working on payment infrastructure. The position is speculative, and the early evidence cuts both ways. Stablecoins have grown to about $315 billion, and the US GENIUS Act of July 2025 requires them to be backed by reserves such as cash and short-term Treasuries, with the law taking effect by January 2027. Almost all are pegged to the dollar, so far spreading its use rather than replacing it.
The reading that best fits all of them: the dollar is still the center of the global monetary system, and that center is being eroded slowly at the edges. The political choices that drive the erosion are mostly American ones rather than Chinese ones, and the technology layer underneath the system is changing faster than the official numbers reveal. A reasonable person can read the data and conclude that the dollar will dominate the next twenty years too, or that the next twenty years will look meaningfully different. Both are defensible. Predictions of imminent collapse and predictions of unchanged dominance are both, however, harder to defend.
What this means for you
For most readers, the dollar's role shows up not in the headlines but in the prices, interest rates, and saving options that touch everyday life. A few practical implications:
If you save in dollars (most American readers)
Your savings are denominated in the world's most widely used currency. The "dollar collapse" predictions of the last twenty years have so far been wrong. The realistic risk to a dollar saver is not collapse but slow loss of purchasing power through inflation, which is a separate and well-documented problem covered on the US Debt page. Holding cash is rarely the right answer; holding a diversified mix of broad stock funds, a modest position in real assets, and some international exposure is.
If you live outside the United States
A strong dollar makes dollar-priced goods more expensive where you live and dollar-denominated debt harder to repay. Many emerging-market countries had a difficult time in 2022 and 2023 for exactly this reason; the dollar's fall in 2025 eased the pressure. If your home country borrows in dollars and earns in something else, your country's economy is exposed to US monetary decisions. Holding some savings in dollars (as foreign-currency deposits, dollar bonds, or US-listed funds) is a reasonable hedge against your local currency weakening; it is also a small bet against the slow-erosion path described above. Both can be true.
If you do business across borders
Outside Europe, where the euro leads, most international trade is still invoiced in dollars, and that has not changed in any practical way. The harder question is sanctions and compliance. An American bank or business now has to do significantly more checking than it did fifteen years ago about whether a counterparty, a country, or a payment violates US restrictions. That cost is real and growing, and it is one of the reasons why some non-US companies have started routing payments around the dollar where they can. If your business has significant exposure to countries that are sanctioned or sanction-adjacent, this is now a lasting risk worth thinking through rather than an edge case.
If you're thinking about long-term wealth preservation
A diversified mix of currencies and asset classes - the kind that passive index funds and target-date retirement funds already provide - is built around the slow-erosion case rather than the collapse case, and that is a reasonable default. Adding a small allocation to gold, international equity, or other-currency exposure is a low-cost hedge against a faster erosion than the base case. Going all-in on any single scenario - all-dollar, all-gold, all-crypto - is a bet rather than a plan, and most plans look better than most bets across full lifetimes.
If you're a citizen of any country
The dollar's role is one of the things your government cannot directly control but does have to react to. Countries that thoughtfully build dollar-equivalent savings and dollar-resilient finance (Singapore, Switzerland, Saudi Arabia, several others) tend to be more robust through monetary shocks than countries that ignore the issue. When you hear domestic political debates about "monetary sovereignty" or "de-dollarization," the practical question is rarely whether to abandon the dollar - it is whether your country has the savings and institutions to handle a dollar shock without crisis. Few countries fully do.
The mechanics behind this
The dollar story sits on top of three deeper mechanisms that show up across this site. If the analysis above depends on ideas you want to understand first, these fundamentals make the conversation more legible:
