Six powers have worn the reserve crown since the early modern era, and each controlled the dominant trade, naval, financial, or industrial network of its age. Portugal controlled routes. Spain controlled bullion. The Dutch controlled finance. France controlled continental power. Britain controlled empire and banking. The United States controlled energy, security, debt markets, and settlement.
The seventh may not be a country.
That isn’t a Bitcoin-maximalist claim. It’s an observation about what the reserve function actually has to do in 2026, and what no single sovereign can credibly do anymore.
The next reserve layer has to clear instantly across borders, settle programmatically, resist unilateral confiscation, and operate above the domestic political cycle of any one state. Those are network properties, not national ones. Issuer dominance becomes network dominance, and the crown migrates from whose currency to which rails.
The mistake people make about this transition is assuming the next reserve must look like the last one. It probably won’t.
What’s Actually Changing
The dollar is not collapsing. Anyone selling that story is selling something else.
As of Q4 2025, the dollar still represented 56.77% of allocated global FX reserves, with the euro at 20.25% and the renminbi at a relatively trivial 1.95%1. The dollar’s lead is not narrowing dramatically; it’s holding. But reserve dominance is not lost when everyone exits.
Reserve dominance isn’t lost when everyone exits. It’s lost when everyone stops needing to add more.
That marginal story is where the signal is. Reuters reported in May that foreign holdings of U.S. Treasuries fell to $9.348 trillion in March 2026, down from a record $9.487 trillion the prior month, with Japan and China leading the reductions2. One month doesn’t make a trend. But the broader pattern — diversification of new flows, growth in non-dollar bilateral settlement, central-bank gold accumulation at multi-decade highs — is consistent with a marginal shift that compounds over years.
The question isn’t “will the dollar collapse.” The question is “will it slowly stop being inevitable.”
The Dollar’s Seven Jobs
The dollar’s dominance isn’t one thing. It’s at least seven simultaneous functions:
- Reserve asset for central banks
- Trade invoice currency
- Commodity pricing unit
- Settlement rail
- Safe collateral base through Treasuries
- Sanctions and enforcement layer
- Global liquidity backstop during crises
A reserve transition doesn’t strip all seven at once. It takes the easiest job first.
That job is settlement.
The dollar is still overwhelmingly used in trade invoicing. Federal Reserve research notes that from 1999 through 2019, the dollar accounted for roughly 96% of trade invoicing in the Americas, 74% in Asia-Pacific, and 79% in the rest of the world, excluding Europe. But settlement is the layer beneath invoicing. A country can continue to price oil in dollars while settling portions of the trade through tokenized deposits, central-bank digital settlement rails, gold-linked claims, commodity-backed credit lines, stablecoins, regional currency corridors, or bilateral netting systems.
That’s the unbundling. The dollar can keep doing some of its jobs even as it loses others… and the first ones to migrate are the plumbing functions where neutrality and programmability genuinely matter more than sovereign brand.
“Project Agorá explores how tokenisation and smart contracts could enhance wholesale cross-border payments, making them faster, more transparent, and more accessible.” — Bank for International Settlements
That’s not maximalist crypto rhetoric. That’s the institutional version of the same thesis.
Phase One: The Dollar Becomes the Protocol
Here’s where the popular framing gets the timeline wrong. The most likely near-term outcome is not “Bitcoin replaces the dollar.” The most likely near-term outcome is that the dollar tries to become the dominant tokenized currency before something else does.
That’s already happening. Regulated stablecoins, tokenized Treasuries, bank-issued deposit tokens, programmable compliance layers, BIS’s Agorá project, U.S.-jurisdiction custody infrastructure — these are dollar-extension mechanisms in their current form. They’re how the dollar absorbs programmable settlement, not how programmable settlement bypasses the dollar.
Think of it as three phases:
- Dollar Dominance 1.0: U.S. banks, Treasuries, SWIFT, Fedwire, sanctions enforcement
- Dollar Dominance 2.0: Stablecoins, tokenized Treasuries, programmable settlement, embedded compliance, regulated tokenization rails
- Post-Dollar Protocol Layer: Multi-asset settlement where dollars are one major asset rather than the system itself
The transition from 1.0 to 2.0 is bullish for U.S. dollar dominance in the short term. Stablecoins are dollar exporters. Tokenized Treasuries are demand vectors for U.S. debt. Programmable compliance is U.S. soft power expressed in code.
If the U.S. executes this transition well — credible regulation, infrastructure investment, balanced fiscal posture — it can extend its monopoly premium by decades.
The question is whether it does.
Phase Two: The Protocol Routes Around the Dollar
The move from Dollar Dominance 2.0 to a post-dollar protocol layer is conditional, not inevitable. It happens if — and largely only if — the U.S. fails at four specific things: defending dollar demand through better digital rails, reducing fiscal recklessness, rebuilding productive capacity, and making Treasuries attractive because the U.S. is credible rather than because alternatives don’t exist.
This is the part the U.S. policy debate hasn’t internalized. The reserve transition isn’t something happening to America. The shape of the next phase depends on what America does in the current one.
The fiscal arithmetic is the constraint that matters most.
Federal interest outlays reached about 3.15% of GDP in 2025, up from 1.49% in 2021. The Congressional Budget Office’s long-term outlook projects net interest reaching 5.4% of GDP by 2055. Those numbers don’t force a transition to Phase Two. But they tighten the runway dramatically, because every additional point of GDP going to debt service is a point not available for the productive investments that would keep the dollar credible as the protocol.
The dollar wins by becoming the protocol. It loses by abdicating that role to fiscal indiscipline.
Phase Two becomes likely when fiscal interest expense crowds out productive capacity faster than productivity gains arrive, when Treasury auctions need progressively higher yields to clear, when sanctions enforcement starts leaking through tokenized rails outside U.S. jurisdiction, and when the marginal incremental settlement flow finds the non-dollar protocol layer simpler and cheaper to use than the dollar one. Those aren’t speculative conditions. They’re observable variables on observable trend lines.
What Compression Feels Like
Households don’t experience reserve transition as monetary architecture. They experience it as compression… the same word that describes the labor-market side of the AI displacement story.
The mechanism is straightforward.
- If the dollar’s purchasing-power premium fades, imports get more expensive.
- If foreign Treasury demand becomes more yield-sensitive, U.S. borrowing costs rise structurally.
- If borrowing costs rise, mortgage rates stay above the post-2008 normal indefinitely.
- If federal interest expense crowds out the budget, fiscal capacity to respond to labor displacement and other shocks shrinks.
The result is a household that pays more for imported goods and energy, holds a mortgage at rates its parents didn’t face, watches federal programs get squeezed, and finds that real wage growth doesn’t catch up because the inflation floor sits at 3–5% rather than 2%3.
This is the same compression the AI inflection produces, arriving from a different direction. AI displaces wage income from the labor side. Reserve transition compresses purchasing power from the monetary side. The household feels both at once.
AI compresses labor. Reserve transition compresses purchasing power. The household feels both at once.
For investors, the framework rhymes with the AI thesis: long pricing power and global revenue, short duration and fragile balance sheets, with hard collateral — gold, productive infrastructure, energy, scarce real assets — as the ballast allocation. The novel addition in the reserve-transition lens is the strategic value of programmable settlement infrastructure: custody, compliance, tokenization rails, identity attestation, the firms that win regardless of whether Phase Two arrives, because they’re the architecture for both phases.
The losers are the inverse profile.
Goldman has already named them as “zombie” balance sheets4, the intersection of high debt loads with weak pricing power, structurally exposed to higher rates and currency erosion at the same time. Add to that list: countries dependent on dollar debt but hostile to dollar governance, banks that can’t modernize cross-border settlement infrastructure, and currencies that only had relevance because the dollar system was willing to intermediate them.
The L-Shape, At the Monetary Scale
The L-shape thesis at the household level says leverage compounds while wages stagnate. The reserve-transition thesis at the geopolitical level says programmable settlement compounds while sovereign currency premium erodes. Those aren’t parallel stories. They’re the same dynamic at different scales — economic power migrating out of human-bounded and nationally-bounded structures and into network-bounded ones.
The people who benefit from the upper branch of the L-shape — those who own assets that compound across the labor displacement — are the same people who benefit from the reserve transition’s winners list, because they’re already positioned on the network layer rather than the labor or sovereign-currency layer. The losers are the same losers: wage-dependent households exposed to import inflation, governments dependent on cheap external financing, businesses without pricing power.
This is why the three essays in this arc fit together. The L-shape diagnosis is the structural picture. The AI earnings inflection is the labor-side timing. This is the monetary-side timing. Three movements of the same composition — and the household sits at the intersection of all three.
The Window You’re Still Inside
The reserve-transition window closes more slowly than the AI inflection window.5 It’s measured in years rather than quarters. But the personal stakes follow the same asymmetry.
By the time the lagging indicators arrive in the form Americans actually notice — mortgage rates that don’t normalize, federal interest expense visibly crowding out programs, sanctions effectiveness declining, gold prices structurally elevated, stablecoin volumes outscaling traditional FX settlement — the adjustments that would have mattered for individual financial security have already gotten harder to make. Hard collateral, non-dollar revenue exposure, productive cash flow, real assets, and optionality on programmable settlement infrastructure are positions available now that get progressively more expensive to acquire later. The reallocation made in 2026 compounds. The reactive one made in 2029 doesn’t.
The L-shape doesn’t have one driver. It has two. The labor side and the monetary side compress on overlapping clocks, and the policy capacity to soften either is being eroded by the same fiscal arithmetic that the reserve transition is itself creating. That’s the runway. It’s the same runway the AI piece described, with a second engine running it down.
The dollar will not collapse. It will be demoted — or it won’t. That outcome is not yet written.
This is the part of the story that the declinist framing gets wrong, and the triumphalist framing gets wrong in the opposite direction. The reserve transition isn’t a fate being inflicted on America by other countries’ choices. It’s a test of whether America still has the institutional discipline to do what built the dollar’s position in the first place: invest in productive capacity, run credible fiscal policy, build the best financial infrastructure in the world, and earn the role rather than inheriting it.
Every previous reserve power lost the crown the same way — not through a sudden defeat but through a slow drift into the assumption that dominance was permanent. Portugal, Spain, the Dutch, the French, the British: each one had its moment of believing the system would carry them regardless of what they did at home. Each one was wrong.
The U.S. is in that moment now. Phase One — the dollar becoming the protocol of programmable settlement6 — is still America’s to win, decisively. The rails, the regulatory infrastructure, the capital markets, the institutional credibility, the dominant currency position are all already in place. What’s required is the political will to defend them: fiscal restraint that makes Treasuries credible on merit, infrastructure investment that makes U.S. tokenization rails the default, productive-capacity rebuilding that gives the dollar something to be backed by beyond inertia, and the discipline to lead the next monetary architecture rather than resist it.
That isn’t a foregone conclusion. But it isn’t a foregone loss either. The seventh reserve doesn’t have to route around the United States. The most likely version of the future — the version supported by current adoption patterns, current capital flows, and current technological lead — is the one where the seventh reserve still has the dollar at its center, just expressed through different rails.
The country that’s currently the sixth still controls how the seventh gets built. That’s the actual stake. Not whether America can prevent a transition — the transition is happening — but whether America stays at the center of the system it transitions into.
Everything depends on what gets done in the window that’s still open.