Nobody Voted for This Empire
Forty percent of the world's dollars rest on an unwritten Fed promise. That promise is now under pressure.
Divide fourteen trillion by five hundred fifty-four billion. The answer, roughly twenty-five, is the working leverage ratio of American monetary power in May 2026.
Every dollar of crisis liquidity the Federal Reserve extended through central bank swap lines at the worst week of the 2008 collapse underwrites about twenty-five dollars of liabilities sitting on bank balance sheets outside the United States today. This is the actual dollar empire. Not the petrodollar arrangement now visibly falling apart in the Strait of Hormuz, not the Treasury auctions that foreign reserve managers attend with declining enthusiasm, not the carrier groups that spent April failing to keep oil flowing out of the Gulf. Not a treaty. Not a vote. Not even a named doctrine. A discretionary backstop, never voted on by Congress, never written into any treaty, never even named in a Fed press release until the morning it gets activated.
Roughly forty per cent of all dollars in existence are created outside the United States, on the ledgers of banks the Federal Reserve does not regulate. The Bank for International Settlements puts the offshore stock at approximately $14tn in liabilities, against $19tn held by the Fed and US commercial banks combined. No other currency comes within an order of magnitude.
Most of the obituaries written for dollar dominance miss this entirely.
When the IMF published its first quarter 2026 COFER data showing dollar reserves at central banks slipping below fifty-seven per cent, the usual think pieces followed. Riyadh diversifying. Delhi diversifying. Moscow gone. Beijing inching upward in renminbi-denominated swap agreements with countries that need cheap commodities and cannot afford diplomatic enemies. The conclusion writes itself, and it is wrong, or at least wrong in the only sense that matters operationally. Work by the New York Fed has already shown that the aggregate reserve decline is driven mostly by idiosyncratic exits rather than systemic flight.
Reserves are sovereign decisions, and sovereign decisions in 2026 are increasingly hostile to Washington. None of that touches the eurodollar pool, because the eurodollar pool is not a sovereign decision. It is a million private decisions made by treasurers in Frankfurt and Singapore and São Paulo who need to fund a six-month invoice and find dollar credit cheaper, deeper, and more liquid than anything available in euros or renminbi. They are not voting for America. They are voting against currency risk. The dollar happens to be on the other end of the trade.
Because those private decisions compound, the offshore pool grows.
Because the pool grows, the Federal Reserve’s implicit obligation to it grows in tandem, even though nobody at the Marriner S. Eccles Building has ever signed a document saying so. This is the part of dollar dominance that almost nobody outside a small circle of monetary economists understands, and it is the only part that actually matters in a year when the petrodollar arrangement is openly disintegrating.
The 2008 swap line extension to the European Central Bank, the Bank of Japan, the Swiss National Bank, and a small group of trusted peers peaked at $554bn within weeks of the Lehman collapse. In March 2020 the figure hit $358bn within days of the pandemic shock. Both interventions are now studied in graduate seminars as triumphs of central bank coordination. Neither was authorized by Congress. Neither involved a vote of any kind by any elected body in any country. Neither came with a published rulebook explaining when the Fed will extend dollars to a foreign counterparty and when it will not.
A standing line of credit from the Federal Reserve to its closest peers, available in effectively unlimited quantity during a crisis, was the load-bearing wall of the global financial system in two of the worst weeks of the past twenty years, and ninety-nine per cent of civilians who use dollar-denominated services every day cannot tell you it exists. The mechanism is the empire. The empire was built by accident, by bankers in the City of London in the 1950s running a remarkable conspiracy of silence, and then quietly endorsed in Washington once it was too large to dismantle. Nobody currently alive at the Federal Reserve was around when the structure took shape, and nobody currently in the Treasury knows how to undo it without producing a banking crisis large enough to swallow the administration that ordered the undoing.
Walk this forward into the calendar of 2026.
The Trump administration is at month seven of an escalating military confrontation with Iran. The Strait of Hormuz has been intermittently closed. Iranian oil moves to Chinese refineries priced in renminbi. Indian state refiners are now settling Iranian barrels through ICICI Bank in yuan, according to Reuters reporting on April 17th. The toll for tankers permitted through the strait was, for at least part of April, denominated in bitcoin. Oil production has dropped in Saudi Arabia and collapsed across Kuwait, Iraq, and the UAE. The petrodollar arrangement that Henry Kissinger sketched into being in the mid-1970s is, by any honest reading of the tape, no longer operative.
The eurodollar arrangement is not the petrodollar arrangement. It has nothing to do with Iran. It has very little to do with Saudi Arabia. It depends on whether, in a crisis, the Federal Reserve still picks up the phone.
After the Treasury imposed a fresh round of secondary sanctions on regional Chinese banks handling Iranian oil receivables, the question of what the Fed would do in a liquidity event involving those institutions moved from academic to operational. The Bank for International Settlements estimates the Chinese banking system holds well over a trillion dollars in offshore dollar liabilities. If a mid-sized Chinese bank suddenly cannot find dollar funding because the New York correspondent network has been instructed by the Treasury to step back, the Fed has a choice. It can extend a swap line to the People’s Bank of China, which does not currently exist. It can lean on the ECB or BoJ to recycle dollars into the gap, which is the same thing routed through one additional counterparty. Or it can decline, watch the bank fail, and let the contagion ripple back through European and Japanese balance sheets that hold the other side of those liabilities.
When the BIS asks the Federal Reserve under what conditions a swap line would be extended to a non-traditional counterparty, the Fed declines to answer in writing. There is no public protocol. There is the discretion of whoever happens to be Fed chair, vice chair for supervision, and the New York Fed president on the morning the call comes in. Three people, two of whom were not confirmed by the Senate to make geopolitical decisions, choosing on roughly an hour’s notice whether to backstop or abandon a meaningful share of the world’s banking system.
Aditi Sahasrabuddhe at Brown University has spent three years documenting the swap line network. Her working paper, circulated through SSRN and updated this winter, makes a quiet but devastating observation. The People’s Bank of China has signed swap agreements with forty countries since 2008. Those agreements have never been tested in a real crisis. The Fed’s lines, by contrast, have been activated twice with full capacity, in 2008 and 2020. The asymmetry has nothing to do with the diplomatic skill of the central bankers involved. When the call goes to Beijing, the country on the other end has to weigh whether the renminbi liquidity is worth whatever security concession the Politburo will eventually request. When the call has gone to Washington for the past twenty years, the answer has been yes, with no conditions, no questionnaire, no follow-up.
If that answer becomes conditional, the eurodollar system stops being free, and once it stops being free it stops being a system. It becomes a series of bilateral favors.
What sits firm on the calendar is the next FOMC meeting beginning June 16th and the Jackson Hole symposium in late August. Neither will mention swap lines. The Fed never does, in normal weather. Both will be parsed for hints about rate paths and balance sheet runoff and the new chair’s policy framework. Almost no one in the room will ask the question that actually matters in May 2026: whether the unwritten promise that has held up forty per cent of the world’s dollars for half a century still holds under an administration that has weaponized every other instrument of American financial power it could find.
The petrodollar arrangement is dying in public, on television, in tanker manifests denominated in bitcoin and yuan. The eurodollar arrangement is something else entirely. It will not die in public.
It will die, if it dies, in a single bad weekend when the Federal Reserve hesitates for forty-eight hours longer than it should, and treasurers in Singapore and Frankfurt and São Paulo all realize at once that they were funding their balance sheets on a courtesy that was withdrawn while they were sleeping.
June 16th. Mark the date.
References:
Statistical Release: Global Liquidity Indicators at end-Q3 2025