On September 12, 2026, leaders from Brazil, Russia, India, China, South Africa, and eight other member nations gathered in New Delhi for the 18th BRICS Summit and adopted the New Delhi Declaration 2026, formally committing the bloc to expand trade and payments in local currencies. The same summit marks the live launch of BRICS Pay, a blockchain-based messaging system designed to let central banks settle cross-border payments without routing through the dollar-denominated SWIFT network. President Trump has already threatened 100% tariffs on any BRICS country that follows through on de-dollarization. Every ingredient for a dramatic “end of dollar dominance” headline is sitting in the same room.
The reserve data tells a much quieter story, and the gap between the two is the actual subject worth analyzing here not because the rhetoric is meaningless, but because the two things rhetoric and reserve data measure are genuinely different, and confusing them leads to bad predictions in both directions.
What’s Actually Launching This Week
Worth taking the ambition seriously before checking it against anything. BRICS Pay integrates several national payment networks that already exist – Brazil’s Pix, China’s CIPS, Russia’s SPFS, India’s UPI into a single messaging layer called the Decentralized Cross-border Messaging System, built to link central bank digital currencies and process up to 20,000 messages per second. A separate initiative, a gold-backed settlement token called The Unit, began pilot testing in early 2026 after an initial 100-unit trial run in October 2025. According to figures circulating around this summit, the BRICS+ bloc, now 11 full members plus a tier of partner countries, together representing something like 40–48% of global GDP by purchasing power, already settles roughly two-thirds of its intra-bloc trade in local currencies rather than dollars.
That’s a real, substantial infrastructure buildout, not vaporware. It deserves to be described accurately rather than dismissed.
What the Reserve Data Actually Shows
Here’s the part that complicates the headline. The dollar’s share of global foreign exchange reserves sits at roughly 58%, according to the Atlantic Council’s Dollar Dominance Monitor is essentially flat, not in freefall. The renminbi, the currency most often cited as the dollar’s most likely successor, held just 1.99% of allocated reserves in the most recent quarter, barely moved from 1.95% three months earlier. If a systemic, coordinated flight from the dollar were genuinely underway among central banks, reserve composition, the slowest-moving, hardest-to-fake data central banks report would be the place it would show up first and clearest. It isn’t showing up there.
There’s an even more specific data point worth sitting with: BRICS isn’t even fully unified behind the loudest version of its own rhetoric. India’s External Affairs Minister, S. Jaishankar, stated plainly in March 2025 that India does not have a policy of replacing the dollar, and has repeatedly distanced New Delhi from the more aggressive de-dollarization framing coming out of Moscow and Beijing. The country hosting this exact summit, in other words, isn’t fully signed on to the version of the story that makes the best headline.
This also isn’t the bloc’s first attempt at exactly this kind of announcement. The 2023 BRICS summit generated a wave of coverage around a proposed common BRICS currency, complete with speculation about gold backing and a basket-of-currencies structure. That specific proposal never materialized into anything central banks actually adopted, and the reserve data from the years since shows no trace of it. The pattern holding across both episodes is the same: a summit produces an ambitious currency announcement, coverage treats it as a structural turning point, and the reserve data a year or two later shows little evidence anything changed. That doesn’t guarantee this year’s announcement fails the same way. BRICS Pay is a more concrete piece of infrastructure than a currency proposal was but it’s a reasonable base rate to weigh against the headlines rather than assuming this time is automatically different.
The Euro Precedent
This isn’t the first time a serious-looking dollar alternative has launched to enormous fanfare, and the last comparable case is worth knowing in detail, because it’s a far stronger test case than anything BRICS has built so far. When the euro launched in 1999, it wasn’t a payment messaging layer bolted onto existing national systems, it was a fully convertible, freely traded single currency backed by one of the largest economic blocs on Earth, replacing the deutsche mark, French franc, and several other established currencies outright. If any alternative was ever positioned to seriously challenge the dollar’s reserve status, it was this one.
The dollar’s reserve share did fall in the following two decades, from roughly 71% in 1999 to about 59% by 2020. But the IMF’s own analysis of that decline found something that undercuts the obvious narrative: the drop was not a shift toward the euro. The euro’s reserve share rose modestly after 2000 but that increase was never sustained, and the currency has sat at a remarkably stable 20% of global reserves for well over a decade; a position it reached fairly early and simply never grew beyond. The IMF’s researchers were explicit that this outcome ran contrary to the widespread expectations at the time that the euro would come to play a substantially larger reserve role. The dollar’s lost share went instead into a wide basket of smaller currencies; the Australian and Canadian dollars, the Swiss franc, eventually the Chinese renminbi, rather than consolidating around any single challenger.
Why Reserve Status and Payment Rails Are Different Layers
The euro case points to the structural reason this gap keeps recurring: reserve currency status and payment infrastructure are related but genuinely separate layers, and building one doesn’t automatically move the other. Reserve composition is a central bank’s long-term store-of-value decision; a slow, conservative portfolio allocation shaped by liquidity depth, legal predictability, and decades of institutional trust, not something that shifts because a faster settlement rail became available. Payment infrastructure, by contrast, is about the mechanics of moving money for a specific transaction. BRICS Pay is squarely a payment-rail innovation: it makes settling a trade between, say, a Brazilian exporter and a Chinese importer faster and less dollar-dependent. It does nothing, by itself, to change what a central bank wants sitting in its reserve account five years from now. The euro proved that even a vastly more credible product on the reserve side of that ledger can plateau far short of dethroning the dollar; a messaging layer aimed at the transactional side of the ledger has an even steeper hill to climb if reserve status is the yardstick.
The Currency Risk That’s Actually Moving Faster
None of that means there’s nothing here for a business to actually watch, it means watching the wrong metric. If the 67% intra-bloc local-currency settlement figure is even roughly accurate, that’s a transactional shift happening on a much faster timeline than reserve composition ever moves, and it’s the layer that actually touches a multinational’s contracts, invoicing, and hedging book. This has precedent too: the “petrodollar” system, where oil exporters priced crude in dollars and recycled the proceeds into dollar assets, took shape in the 1970s and persisted as the overwhelming default for decades essentially independent of what was happening to the dollar’s reserve share in the same period, including through the entire stretch when the euro was supposedly going to challenge it. Invoicing currency and reserve currency status are correlated but genuinely decoupled variables, and they can move at very different speeds for very different reasons, in either direction.
The practical implication for currency risk management is straightforward, and it’s the same discipline this blog has argued for when a single concentrated statistic gets asked to carry a bigger story than it can support: a treasury team watching only reserve-share headlines and concluding “the dollar’s fine, nothing to hedge differently” could still get blindsided by a supplier or customer inside a BRICS-heavy supply chain increasingly invoicing in yuan, rupees, reais, or rubles well before any of that shows up in IMF reserve data. The two data series this post has been comparing aren’t actually in conflict, they’re tracking different things, moving at different speeds, and a business exposed to BRICS trade flows needs to watch the faster one directly rather than inferring it from the slower one.
What the Data Actually Predicts
Two specific, checkable calls follow from putting the rhetoric-versus-data comparison and the currency-risk layer together.
On the reserve-composition side: expect the next several IMF COFER releases the first meaningful test being the Q4 2026 data due in early 2027, to show at most marginal movement in dollar and renminbi reserve shares attributable to this week’s summit. The Atlantic Council’s own house view already anticipates subdued near-term movement, and the euro precedent suggests that even a far more credible reserve alternative took years to plateau rather than months to surge. If COFER data shows dollar share holding in the high-50s and renminbi share still hovering near 2% by mid-2027, that’s this prediction confirmed; a sudden multi-point renminbi jump would be the signal that this time is genuinely different.
On the transactional side: expect intra-BRICS local-currency settlement share to keep climbing measurably faster than reserve data over the next few years, the same way dollar invoicing in commodities decoupled from dollar reserve trends for decades. That’s the metric worth building into actual treasury planning not because it threatens the dollar’s throne, but because it’s the one that shows up in a purchase order before it ever shows up in a central bank’s balance sheet.
The Ground Floor Take
Both readings of this week’s summit are true, and they’re not actually in tension once the layers are separated. The dollar’s reserve dominance isn’t under near-term threat from anything launched in New Delhi this week the data doesn’t show it, and the euro’s quarter-century of plateauing at 20% is a reminder that even a far more credible challenger struggled to move that specific needle. But dismissing the summit as pure theater misses the layer that’s actually moving: a growing share of real, physical trade between BRICS members is settling outside the dollar right now, on a timeline measured in years rather than decades, and that’s the exposure that shows up in a contract long before it shows up in a reserve report. The mistake this week’s coverage risks, in either direction, is treating one headline number as the whole picture the same trap worth avoiding whether the concentrated story is a handful of banks’ earnings or a single reserve-share statistic getting asked to explain more than it actually can.
