The dollar's quiet erosion, and what the central banks are quietly saying
For the first time since 2023, more reserve managers plan to trim dollar holdings than add to them. That signal deserves more weight than the equity tape currently gives it.
The most consequential number floating across markets on 9 July 2026 isn't an inflation print or a dot on a chart. It's a tally of intentions: for the first time since the GPI series began recording reserve managers' long-term plans in 2023, more central banks say they intend to decrease dollar holdings over the next ten years than say they intend to increase them. The signal, posted in the early hours of 9 July, is a slow-moving crack in a wall the rest of the tape still pretends is solid.
The equity market, in the meantime, is busy being expensive. Bank of America's screen flagged the S&P 500 as "statistically expensive on 17 of 20 metrics," trading rich versus tech-bubble benchmarks on eight of them, and urged investors to take profits in July. The S&P 500 closed at a level where almost every conventional valuation gauge is in the top decile of its post-2000 range. None of this is hidden. It is, in fact, conspicuously advertised. The market simply doesn't care, because the marginal buyer is leveraged, indexed, and momentum-driven rather than valuation-driven.
Put those two facts side by side and the divergence starts to make sense. The equity market is pricing in American exceptionalism: AI capex, durable growth, a soft landing that refuses to die. The reserve managers are pricing in something else. They are pricing in a world in which the United States increasingly weaponises the dollar's plumbing — secondary sanctions, correspondent-banking cutoffs, asset freezes — and in which that weapon starts to cost the weaponiser. The marginal reserve manager, after all, is a treasurer whose job is to keep a national balance sheet investable across a multi-decade horizon, not to ride a single equity cycle.
The Fed minutes released earlier this week sharpened the policy backdrop. Rate cuts are losing internal support as inflation persists, which means the carry advantage of holding Treasuries is narrowing just as the political premium for holding them is widening. A central bank looking at a 10-year horizon sees the prospect of structurally higher Treasury yields and structurally less willingness among foreign holders to absorb the supply. Some of those reserve managers have already started to act: gold buying by emerging-market central banks has run well above the pre-2022 average, and the renminbi's share of cross-border settlements continues to creep higher. The GPI datapoint is the survey confirmation of what the gold and the settlement numbers have been telling us for two years.
The structural reading is straightforward. The incumbent order is ceding ground to a successor arrangement — not because the dollar is collapsing, but because the cost of holding it is rising. There is no single moment at which reserve currency status flips; it bleeds. A few basis points of gold allocation here, a bilateral local-currency swap line there, a few percentage points of cross-border trade settled outside the dollar. The aggregate is what matters, and the aggregate is now pointed in one direction.
What remains genuinely uncertain is the pace. The dollar remains dominant by every conventional metric — something like 58% of allocated reserves, the majority of cross-border invoicing, the monopoly on the most liquid safe asset. Reserve managers are a cautious, herd-like population; they typically move in the same direction, with a lag. The GPI reading is a signal of intent, not a record of action. The counter-argument, taken seriously, is that U.S. capital markets are simply too deep and too liquid to displace in any horizon a reserve manager actually plans for, and that talk of de-dollarisation has been overhyped for two decades. There is something to that. The dollar's network effects are real.
The counter to the counter is that network effects erode asymmetrically when the network operator starts to fine the users. The Treasury and OFAC actions of the last several years have given every treasurer in the Global South a vivid, concrete reason to diversify. None of those treasurers are going to dump dollars next quarter. All of them are going to keep adding a little less, and tilting a little more toward gold, euros, renminbi, and yes, even crypto-settled bilateral arrangements. That is exactly what the GPI tally is now capturing.
The stakes are concrete. If the trend continues, U.S. fiscal dominance — the quiet subsidy the rest of the world pays by parking savings in Treasuries at compressed yields — narrows. That means higher long-end yields at any given debt-to-GDP ratio, which means harder politics in Washington. It also means a quieter transfer of influence to whatever settlement systems and reserve assets absorb the marginal dollar. Whoever builds those rails — Beijing, the Gulf states, the BIS, a constellation of bilateral arrangements — will own the next era's plumbing.
What the sources do not yet specify is the magnitude of the planned decrease, or how concentrated it is among the largest holders. The GPI series is intention, not execution, and the gap between the two is sometimes a decade wide. The prudent read is also the unglamorous one: the dollar's reign ends not with a crisis but with a long, patient rebalancing, and the first hard evidence of that rebalancing is sitting in a survey most equity desks are not yet reading.
This publication treats reserve-manager surveys as a slow indicator with disproportionate signal, and pairs them with the bond and gold tape rather than the equity tape. The wire coverage of the GPI reading remains thin; the equity market is currently priced for a world the central banks are quietly walking away from.
Wire provenance
This editorial synthesis draws on the following public wire/social posts:
- https://x.com/unusual_whales/status/2026-07-09-dollar-gpi
- https://t.me/CryptoBriefing
- https://t.me/CryptoBriefing