marker 03

Trade Settlement Share Will Decouple from Reserve Share

status: on-track
reading dated: July 30, 2026 link permalink
published: October 6, 2026
next reading: October 30, 2026
reading id: M3-2026Q3

from Appendix G

The lens suggests we should expect the historical correlation between a currency's trade settlement share and its global reserve share to weaken over the coming decade as digital payment infrastructure enables trade settlement in currencies that are not held as reserves. Specifically, the Chinese yuan's trade settlement role should continue to grow in specific corridors faster than its reserve share, and several regional currencies should achieve meaningful trade settlement roles without corresponding reserve accumulation.

The diagnostic logic: Digital payment rails reduce the conversion friction that historically required traders to hold reserves in settlement currencies. As conversion friction drops for non-dollar currencies, the link between "settling trade in X" and "holding reserves of X" loosens. This decoupling is a structural shift in how network dominance operates—settlement becomes a technology problem rather than a reserve problem.

what to watch

The ratio of trade settlement share to reserve share for the yuan, rupee, real, and other regional currencies. Directional divergence—settlement use rising faster than reserve accumulation—would confirm the decoupling pattern.

current reading

The IMF's COFER release of July 1 put the renminbi at 1.99% of allocated global reserves for Q1 2026—up from 1.95% a quarter earlier, still below its 2022 peak near 2.8%. Over the same half-year, CIPS settled ¥100.9 trillion ($14.7 trillion) in cross-border renminbi payments, annualizing to roughly ¥202 trillion, about 12% above 2025's full-year total. On SWIFT's measure the renminbi carried 8.00% of global trade finance in June, second to the dollar at 81.16% and ahead of the euro at 5.59%—though that series is volatile, and May's print was 5.51%, below the euro. The renminbi's share of SWIFT payments overall held between roughly 2.6% and 3.1% through the first half, fifth or sixth place depending on the month. Settlement and trade-finance use of the renminbi continue to run at multiples of its reserve weight. The two series have not reconverged.

CIPS held at elevated levels without accelerating. The system processed ¥50.1 trillion ($7.3 trillion) in Q2 2026, flat against Q1's ¥50.9 trillion; June's ¥18.2 trillion ($2.67 trillion) was the second-highest month of the half after the March spike. One correction to the Q2 reading belongs here. 2025 did not reach the $30 trillion that reading projected: CIPS settled ¥180.15 trillion ($25.55 trillion) in 2025, up 2.7% from 2024—a sharp deceleration from 2024's 43% growth, and the data was available when the Q2 reading was written. The 2026 run-rate is a re-acceleration from that flat year, not a continuation of 2024's pace. CIPS counted 1,766 participants across 124 countries at end-2025, up from 1,629 across 119 a year earlier.

The reserve side moved slightly against the trend this quarter, and the reading should say so. COFER's Q1 2026 data showed the dollar's share of allocated reserves rising to 57.13% from 56.42%, with roughly half the increase attributable to exchange-rate valuation. The euro fell to 20.03%. The residual "other currencies" category declined to 6.18% from 6.25%—the first decrease after seven consecutive quarterly increases, and a dent in one of the trends the Q2 reading cited. (That reading gave 6.13% for Q4 2025; the IMF's current series shows 6.25%.) The World Gold Council's June survey found 74% of reserve managers expect the dollar's share of global reserves to decline over five years—expectation, not yet allocation. One quarter does not reverse a multi-year pattern, and half the dollar's gain was mechanical. But the dollar's reserve share did not fall this quarter, and the record should show it.

The marker's claim concerns the relationship between the two series, not the direction of either one: trade settlement share and reserve share no longer move together. That relationship held. CIPS ran near $30 trillion annualized while renminbi reserves stayed under 2%; the gap did not narrow. Status reads as on-track. What would call this into question: trade settlement share and reserve share resuming their historical correlation; CIPS volume growth flatlining or reversing; a sustained reversal in the residual "other currencies" reserve trend; or the emergence of a credible non-Chinese settlement rail absorbing renminbi-denominated flows. The second and third conditions each registered a single quarter's data point—CIPS flat against Q1, "other currencies" down seven basis points. Neither is yet sustained. The first and fourth have not appeared.

sources for this reading

standing sources

past readings

May 25, 2026 link permalink on-track read in full collapse

CIPS volumes up 43% YoY to $24.5T (2024); RMB SWIFT share declined to 2.13% (Jan 2026) while CIPS hit record $178.5B single-day in April 2026; RMB COFER reserve share declined to 1.95% Q4 2025; settlement migrating off SWIFT to CIPS while reserve share falls within legacy framework.

Trade settlement and reserve share for the yuan have moved in opposite directions over the past year, by significant margins. China's Cross-Border Interbank Payment System (CIPS) processed approximately $24.5 trillion in cross-border RMB transactions in 2024, up 43% year-over-year, with 2025 on pace to reach roughly $30 trillion. Average daily CIPS settlement value rose from 619 billion yuan in February 2026 to 920 billion yuan in March 2026—a 50% monthly increase—and hit a single-day record of 1.22 trillion yuan ($178.5 billion) in early April. Over the same period, RMB share of SWIFT international payments declined from a December 2024 peak of 4.6% (fourth place globally) to 2.13% in January 2026 (sixth place). Yuan-denominated trade settlement is not in retreat. It is migrating off the Western messaging system that historically measured it.
The reserve side moved in the opposite direction. RMB share of allocated official reserves declined from 2.18% in late 2024 to 1.95% by Q4 2025—a roughly 10% relative drop in twelve months. The COFER residual "other currencies" category, which captures reserves held in currencies the IMF does not separately track, has more than doubled since 2021 and reached 6.13% in Q4 2025. Two patterns appear at once: RMB trade settlement growing at double-digit rates outside legacy rails, RMB reserve share declining within them. The historical correlation between trade settlement share and reserve share—the link the marker predicted would weaken—is no longer just weakening. It is breaking in opposite directions for the same currency.
The mechanism is partly geopolitical and partly infrastructural. CIPS now includes direct participants from the Middle East, Africa, Central Asia, and Singapore's offshore RMB center; Beijing updated CIPS operating rules in early 2026 to support multi-currency settlement, positioning it as a platform rather than a yuan-only rail. The March 2026 CIPS volume spike coincided with the Iran conflict that began that February, though the Atlantic Council has cautioned the data shows broader growth in renminbi settlement capacity rather than direct proof of Iran-linked flows. Bilateral local-currency settlement arrangements continue to expand across the Global South, which now accounts for 44% of China's exports. None of this is a coordinated displacement of the dollar—the dollar still settles 58% of SWIFT international payments and dominates reserves at 56.77%. (The IMF later revised this figure to 56.42%; see the reading dated July 30, 2026.) But the architecture of settlement is fragmenting: more rails, more bilateral arrangements, more currencies that don't fit the legacy COFER framework.
The framework's central claim about Marker 3—that digital payment rails would reduce the conversion friction that historically required reserve holdings to enable settlement—is being borne out, with the additional wrinkle that the rails themselves are fragmenting along geopolitical lines. The yuan's global SWIFT share has declined, but its settlement use in specific corridors and through alternative infrastructure has grown substantially, exactly the pattern Appendix G named. Status reads as on-track. What would call this into question: trade settlement share and reserve share resuming their historical correlation, CIPS volume growth flatlining or reversing, a sustained reversal in the residual "other currencies" reserve trend, or the emergence of a credible non-Chinese alternative settlement rail that absorbs RMB-denominated trade flows. The opposite is happening on the first three; the fourth has not appeared.

note added October 23, 2026—see the corrections page
sources for this reading

what would call this into question—from Appendix G

Trade settlement share and reserve share continuing to move together as they historically have. If currencies that gain trade settlement role also gain proportional reserve share, the link the lens predicts will weaken has held instead.