Dedollarization

How far the world has moved away from USD dominance in reserves, trade, and safe-haven demand.

42
/ 100
Early shift

A score of 0 means the US dollar is fully dominant in global reserves, trade, and safe-haven demand. 100 means a post-dollar multipolar system. Components are weighted from IMF COFER (currency composition of reserves, USD at 56.8%), FRED (trade-weighted USD indexes, foreign Treasury holdings, gold price). Higher gold prices and rising CNY reserve share push the score up. A strong DXY and steady foreign demand for US Treasuries push it down.

What central banks hold

Currency composition of allocated foreign exchange reserves, plus the separate gold allocation. Dedollarization is visible here: USD was 72% in 2002, now 56.8%.

Allocated FX reserves (COFER)
USD 56.8%
EUR 20.2%
JPY/GBP 14.9%
CNY 2.0%
Other 6.1%
Gold share of reserves
~18%

Central banks held roughly 18% of reserves in gold as of late 2024, up from ~10% in 2015. COFER tracks only currency reserves, so gold shows up here as a separate allocation.

Source: World Gold Council (no programmatic API, manually refreshed).
USD decline since 2002
−15.2pp

The USD's reserve share peaked at 72.0% in 2002. Today it's 56.8%, a 15.2 percentage-point decline over two decades. Slow but structural.

Source: IMF COFER (quarterly, 6-month lag).

Reserve composition, year by year

Drag the slider from 1970 to 2025. The currency bars rebalance live. The gold bar shows gold as a percentage of total reserves (including gold); the currency bars show each currency's share of FX reserves alone.

Year 2025
19701985200020152025
Share of allocated FX reserves
USD
56.8%
EUR
20.3%
CNY
1.9%
Other
21.0%
Gold share of total reserves (different denominator)
Gold
19.0%
2024 · Central bank gold buying record

Central banks buy over 1,000 tons of gold for third consecutive year. Highest sustained rate since the 1960s.

The USD's reserve share has dropped from roughly 72% at its modern peak (2002) to 57% today. Most of that slack didn't go to the euro or renminbi. It went to "Other" (AUD, CAD, KRW, SGD) and to gold. Central banks don't announce their moves. They show up in the quarterly COFER release months later.

Sources: IMF COFER · gold share derived from World Gold Council vs IMF IFS total reserves.

The post-Nixon arc

On August 15, 1971, Nixon closed the gold window and ended dollar convertibility into gold. The dollar became fiat and took on a new role as the world's reserve currency by convention rather than by backing. The charts below trace what happened to USD dominance and to gold since that day.

USD vs gold share of global reserves

Both as a percent of total central bank reserves (FX + gold + SDRs). USD derived from IMF COFER adjusted for the gold share. Gold from World Gold Council and IMF IFS.

0%15%30%45%60%75%90%197019801990200020102020Nixon shockRussia sanctions46%19%
USD % of reserves reserves
Gold % of reserves reserves

Gold price, USD per ounce

London fix, nominal USD. Before 1971 the price was pegged at $35.

$0$500$1.0k$1.5k$2.0k$2.5k$3.0k1960197019801990200020102020Gold window closesGFCCB buying surge$2.4k

Sanctions ↔ reserve share

Each rose marker is a sanction or dedollarization milestone, placed on the USD reserve share curve at the year it happened. Click any marker to see what happened and how the share responded.

50%60%70%80%90%197019852000201520251971197419992014201520162022202320232024

Click a marker above to read the event. The 2014 Russia sanctions and 2022 reserve freeze are the two events that visibly bend the curve.

Correlation isn't causation. USD share was already drifting down from its 2002 peak before the 2014 Russia sanctions. But the sanctions accelerated the diversification by signaling to every non-US central bank that dollar reserves carry legal risk. The 2022 freeze of Russian reserves was the loudest version of that signal so far.

Sources: USD share from IMF COFER · event timeline compiled from Treasury OFAC, IMF, and contemporary press reporting.

Foreign share of publicly-held US Treasuries, 2000 to today

% of publicly-held marketable debt

In 2008 foreigners held 56% of publicly-held US Treasury debt. The share has drifted down to 27%, a 29-point drop. The pivot points are the 2008-09 GFC (Fed buying displaced foreign demand), 2014 (China stopped reserve accumulation), and 2022 (the freeze of Russian reserves made every non-aligned central bank reprice the sovereign risk of holding US debt). Foreigners are not selling in size. They are not buying the new issuance. The Treasury is being financed increasingly at home, by the Fed, banks, money funds, and households, at prices the domestic market will accept.

10%20%30%40%50%60%Pre-GFC peakPost-GFC plateauSanctions inflectionNow20002008201420222026
Pre-GFC peak: 2008: 56%. Foreign central banks recycled trade surpluses into Treasuries through the 2000s. Bretton Woods II was running at full tilt.
Post-GFC plateau: 2014: 46%. Quantitative easing put the Fed onto the buyer-of-last-resort role. China stopped accumulating. Foreign share started its multi-year drift down.
Sanctions inflection: 2022: 30%. Freezing Russian reserves made every non-allied central bank rethink the safety of US debt. Gold buying accelerated. Saudi Arabia tested non-dollar oil settlement.
Now: 2026 Q1: 27%. Foreigners now fund roughly a quarter of US borrowing. The remaining three-quarters comes from the Fed, banks, money market funds, and households. The marginal buyer at the long end is increasingly domestic and price-sensitive.
2008 PEAK
56% · 2008
CURRENT
27% · 2026
DROP
−29 pp
Foreign demand is falling against a Treasury market that needs to grow. US federal debt is running near 125% of GDP, modestly above the 1946 WWII peak, and CBO baseline projections have it climbing through the 2030s. See the debt/GDP chart on /fed for the fiscal-capacity backdrop.
Source: Treasury TIC foreign holdings of US Treasury securities, divided by publicly-held marketable Treasury debt from the Treasury Monthly Statement. Year-end snapshots; 2026 is Q1.

The petrodollar

The "oil priced in dollars" line on the timeline above is the short-form for a specific 1974 arrangement: Saudi oil revenues recycled into US Treasuries through a confidential add-on auction, in exchange for security guarantees. That mechanism produced 50 years of structural foreign demand for USD and US debt, and it is the single largest source of the network effect that makes dedollarization slow.

What is changing now is not just sanctions politics. Three pressures are working at once. Oil's long-run terminal value is being repriced by the energy transition (see the producer side on the climate page). The largest producers are being pulled into a non-USD-aligned political bloc through BRICS+. And as of April 28 2026, the cartel itself is fragmenting: the UAE has announced it will leave OPEC and OPEC+ effective May 1, ending six decades of membership and removing a binding production constraint on the rest of the cartel. The asset class that has anchored the USD for half a century is now on a declining demand curve, under a weakening price-discipline mechanism, with its largest producers shifting alignment. Profits and buybacks at Western majors over 2022-23 read as terminal-value extraction by an industry whose geopolitical anchor is decaying on three axes simultaneously.

Saudi Arabia US Treasury holdings, 2012-2024

USD billions, end-of-December (2020 plotted at the February peak, the all-time high). The Saudis stopped expanding the recycling channel after 2020, with the difference reallocated into gold, real assets, and CNY-denominated instruments. Source: US Treasury TIC Major Foreign Holders archive.

$0B$50B$100B$150B2012201420162018202020222024$138B
1974

US-Saudi commission + confidential Treasury recycling

June 8 1974: Kissinger and Prince Fahd sign the US-Saudi Joint Commission on Economic Cooperation in Washington. December 1974: Treasury Undersecretary Jack Bennett finalizes a confidential add-on arrangement letting Saudi Arabia buy US Treasuries outside normal auctions. Saudi holdings are grouped under "oil exporters" in TIC data and not separately disclosed until a 2016 Bloomberg FOIA. The combination of dollar-priced oil and recycled surplus into Treasuries becomes the structural foundation of the post-Bretton-Woods USD.

2002

USD reserve-share peak

IMF COFER: USD share of global central-bank reserves peaks at ~72% in Q3 2001 to 2002. Every drift down since (currently ~57%) is happening from this baseline.

2018

Shanghai INE launches yuan-denominated crude oil futures

March 26 2018: the Shanghai International Energy Exchange (INE) launches yuan-denominated crude oil futures (contract code SC), China's first commodity future open to foreign investors. Initial volume reaches 14-20% of front-month global crude futures within the first year. The contract is the first credible non-USD oil benchmark and the technical infrastructure for any subsequent yuan oil-settlement deals.

2020

Saudi US Treasury holdings peak

February 2020: Saudi Arabia holds $184.4B in US Treasuries (Treasury TIC), the highest on record. By September 2024 the figure is ~$144B, with the difference reallocated into gold, real assets, and CNY-denominated instruments. The recycling channel is no longer expanding.

2022

Saudi-China yuan oil pricing under discussion

March 15 2022 (WSJ): Saudi Arabia is in active discussions with Beijing to price some of its oil sales to China in renminbi. The reporting follows the Feb 2022 freezing of Russian central-bank reserves, which restated holding USD as a political decision for every non-aligned reserve manager. December 2022: Xi Jinping in Riyadh formally pledges use of the Shanghai Petroleum & Natural Gas Exchange (SHPGX) for renminbi settlement of oil and gas trade.

2023

PBOC-SAMA renminbi-riyal currency swap

November 2023: the People's Bank of China and the Saudi Central Bank (SAMA) sign a CNY 50bn / SAR 26bn (~$7B) bilateral currency swap, providing the operational liquidity for direct CNY-SAR trade settlement without USD intermediation. It is the first such swap between the two central banks.

2024

BRICS+ expansion brings major oil producers inside

January 1 2024: Iran, the UAE, Egypt, and Ethiopia formally join BRICS+. Saudi Arabia received an invitation but, as of late 2024, has not formally accepted, participating as an invited partner. Iran and the UAE alone account for ~6% of world oil production. The expansion places more of OPEC+ output under a non-USD-aligned political bloc actively building parallel settlement infrastructure (CIPS, mBridge, BRICS Pay).

2026

UAE exits OPEC and OPEC+

April 28 2026: the UAE announces it will leave OPEC and the broader OPEC+ supergroup effective May 1, ending six decades of membership. The UAE was OPEC's third-largest producer at ~12% of cartel supply before the Iran war. The Energy Minister framed the move as a "pure policy change" toward "accelerated production" rather than a rejection of Saudi-led discipline, but the years of friction over UAE production caps are the proximate cause. Three structural consequences fall out of this. (1) OPEC's ability to coordinate price discipline weakens; the third-largest producer is now an independent maximizer, which removes a binding constraint on the rest of the cartel. (2) Higher UAE production at the margin is bearish for oil prices, which compresses Western-major profits and accelerates transition economics on the demand side. (3) The cartel that has set the price of the asset class anchoring the USD for fifty years is fragmenting at the same moment its largest producers are formally inside (Iran, UAE) or invited into (Saudi Arabia) a non-USD-aligned political bloc. Each of the three consequences pulls in the same direction the timeline above already points.

Events that moved the needle

Dedollarization doesn't move in a straight line. It moves in reaction to specific events that reminded non-US central banks that holding dollars is a political decision, not just a financial one. Every line in the timeline below nudged the trend.

1971

Nixon closes the gold window

USD becomes fully fiat. Bretton Woods ends. The modern reserve-currency regime begins.

1974

Petrodollar deal

Saudi Arabia agrees to price oil exclusively in USD in exchange for US security guarantees. Locks in USD demand for 50 years.

1999

Euro launched

Europe creates the first credible alternative reserve currency. Eventually takes ~20% of reserves.

2014

Russia sanctions after Crimea

Moscow begins selling US Treasuries and diversifying reserves. The first major "dedollar" move by a large reserve holder.

2015

CIPS launches

China opens Cross-Border Interbank Payment System as a yuan-denominated alternative to SWIFT.

2016

CNY joins IMF SDR basket

Renminbi added to the Special Drawing Rights basket alongside USD, EUR, JPY, GBP. Official reserve currency status.

2022

Russian reserves frozen

$300B of Russian central bank reserves frozen after Ukraine invasion. Every non-US central bank asks: could this happen to us?

2023

BRICS+ expansion

Saudi Arabia, UAE, Iran, Egypt, Ethiopia invited to join BRICS. Explicit talk of alternative settlement currencies.

2023

Saudi-China yuan oil deal

Saudi Aramco executes first yuan-settled oil sale to China. Petrodollar cracks.

2024

Central bank gold buying record

Central banks buy over 1,000 tons of gold for third consecutive year. Highest sustained rate since the 1960s.

BRICS+ vs G7 at scale

Talk of a BRICS+ reserve currency is easier to dismiss when you picture the old bloc of Brazil, Russia, India, China, South Africa. It's harder to dismiss after the 2024 expansion added Iran, UAE, Egypt, Ethiopia, and Indonesia. By population and oil production, BRICS+ already outscales the G7. By GDP it's still smaller but growing faster.

BRICS+

post-2024
Members 10

Brazil, Russia, India, China, South Africa, Iran, UAE, Egypt, Ethiopia, Indonesia

Population 3.6B

45% of world population

Nominal GDP $30T

28% of world GDP

Share of world trade ~23%

Up from 11% in 2000

Oil production 40%

OPEC+ overlap with BRICS makes this bloc-controlled

G7

unchanged since 1997
Members 7

US, UK, France, Germany, Italy, Japan, Canada

Population 780M

10% of world population

Nominal GDP $48T

43% of world GDP

Share of world trade ~30%

Down from 45% in 2000

Oil production 18%

Almost entirely US plus Canada

Share of world nominal GDP

IMF World Economic Outlook. G7 has been shrinking as a share of world GDP for 25 years. BRICS+ has been catching up. The lines cross sometime this decade.

0%10%20%30%40%50%1990199520002005201020152020BRICS+ expansion30%%28%%
G7
BRICS+

SWIFT vs CIPS

The plumbing of international finance runs on messaging systems. SWIFT, based in Belgium, handles most of it. China launched CIPS in 2015 to route yuan settlement without needing SWIFT's cooperation. CIPS is still tiny in absolute terms but its growth rate is what matters. On sanctioned corridors (Russia-China, Iran-China) CIPS is now the primary plumbing.

SWIFT

+5% YoY
~45M
messages / day

Global interbank messaging since 1973. Still handles most cross-border flows.

CIPS

+25-40% YoY
~40K
messages / day

China's yuan-denominated system. Small in absolute terms but growing fast, especially on BRICS trade corridors.

SWIFT vs CIPS daily message volume

Log scale (each gridline is 10×) so both systems can fit on one axis. SWIFT sits around 50M/day. CIPS is much smaller but compounding fast.

1001K10K100K1M10M100M2016201820202022202450M40K
SWIFT
CIPS

CIPS is ~0.09% of SWIFT by volume. At its current growth rate (~30% YoY vs 5% for SWIFT), it would take about 25 years to catch up. Most analysts don't think it will. The more likely scenario is a bifurcated system: SWIFT for the West, CIPS + bilateral swap lines for a BRICS+ bloc. The USD's network effect shrinks without being replaced.

Sources: SWIFT · CIPS volume figures per their annual reports.

What moves the score

Pushing up concerning
USD share of global FX reserves 48
Gold price (export index) 92
Other currencies share (AUD/CAD/KRW/etc) 51
Holding down stabilizing
Foreign holdings of US Treasuries 25
CNY share of global FX reserves 20
USD vs emerging-market currencies 33

Historical context

The USD's share of global central bank reserves peaked near 72% in 2002. After the 2008 crisis, the 2014 sanctions on Russia, and especially the 2022 freezing of Russian reserves, countries began diversifying. That share has now drifted down to 56.8%. The euro took some of the slack in the 2000s but has plateaued around 20.0%. The renminbi gained rapidly after 2016 when the IMF added it to the SDR basket, then stalled around 2.0% as China's growth slowed and capital controls limited convertibility. The real shift has been to non-traditional currencies (AUD, CAD, KRW, SGD) and gold. Central banks bought more gold in 2022-2024 than in any comparable period since the 1960s. Gold is up roughly 3x since 2000. Dedollarization is structural and slow. The USD's network effects (oil priced in dollars, SWIFT, Treasury market depth) make it sticky. Sanctions weaponization and US fiscal trajectory keep the pressure on.

Component breakdown

USD share of global FX reserves
reserves · 30% wt
56.8
score 48

What % of all central bank reserves are held in USD. Was 72% in 2002. Lower = more dedollarization.

CNY share of global FX reserves
reserves · 15% wt
1.95
score 20

Chinese renminbi's share of central bank reserves. Barely existed before 2016. Rising = progress toward multipolar.

Other currencies share (AUD/CAD/KRW/etc)
reserves · 10% wt
6.13
score 51

Share held in smaller reserve currencies like AUD, CAD, KRW. Rising = diversification away from USD.

USD broad trade-weighted index
fx · 15% wt
118.4
score 33

How strong USD is vs a basket of trading partner currencies. Higher = stronger USD = less dedollarization pressure.

USD vs emerging-market currencies
fx · 10% wt
128.5
score 33

USD strength vs emerging markets specifically. EMs are where dedollarization rhetoric is loudest.

Foreign holdings of US Treasuries
debt · 10% wt
9248
score 25

How many US Treasuries foreigners hold. Rising = they still want USD assets. Falling would be a major signal.

Gold price (export index)
flight · 10% wt
181.6
score 92

Gold price relative to its historical range. Central banks and investors buying gold = flight from fiat / USD.