The Queue Nobody Talks About
It is sometime after midnight in the treasurer's office of a mid-sized central bank. The offshore dollar funding rate is moving in ways her models did not anticipate, ticking upward in increments that stopped being noise about forty minutes ago. She knows her institution holds a swap line agreement with the Federal Reserve. What she may not know, and what almost nobody outside the relevant policy rooms bothers to ask, is whether that line is the kind that gets honored in full, the kind that gets rationed down to whatever remains of a preset ceiling, or the kind that quietly stopped being available before she even reached for the phone. The distinction is not technical fine print. It is the difference between a genuine backstop and a politely worded ceiling.
The answer lives inside the architecture of the swap line itself. That architecture is not uniform.
Two Families, Very Different Guarantees
The Federal Reserve's network of currency swap lines divides, structurally, into two distinct families. The first is the standing, uncapped arrangements: agreements maintained with the European Central Bank, the Bank of England, the Bank of Japan, the Bank of Canada, and the Swiss National Bank. These five are sometimes called the C6 group, the Fed being the sixth member. Under these arrangements there is no preset ceiling on the volume of dollars that can be drawn. The ECB, in principle, can request as many dollars as its domestic banking system demands, and the Fed will supply them against an equivalent euro deposit. The credit risk to the Fed is minimal because it holds the foreign currency as collateral at the same exchange rate used to open the swap, so currency movements cannot erode its position.
The second family is the limited, capped arrangements. These were extended, at various points, to central banks including the Reserve Bank of Australia, the Banco do Brasil, the Bank of Korea, the Monetary Authority of Singapore, and others. The caps matter enormously, and it is worth being blunt about why: a line capped at thirty billion dollars is not merely smaller than an uncapped line. It is categorically different. Once that ceiling is reached, the central bank on the other end cannot draw further, regardless of how severe the dollar shortage in its jurisdiction becomes. The cap is the mechanism by which the Fed implicitly ranks its counterparties, and anyone who describes these two families as equivalent is either uninformed or being diplomatic to the point of dishonesty.
So the first way a central bank loses dollar access is simply arithmetic. It exhausts its cap while the stress is still building.
The Mechanics of a Draw, Step by Step
To understand where the system can fail, it helps to walk through a single transaction in some detail. Say the Bank of Korea's cap sits at sixty billion dollars. Korean commercial banks are borrowing heavily in dollars offshore to fund trade finance and cross-border lending. Dollar funding markets seize up. The Bank of Korea activates its swap line.
It deposits Korean won with the Federal Reserve at the current spot exchange rate. The Fed simultaneously deposits an equivalent dollar amount with the Bank of Korea. The Bank of Korea then lends those dollars to Korean commercial banks through a repo-style auction, taking high-quality collateral in return. At the end of the swap term, usually seven days or eighty-four days, the transaction reverses: the Bank of Korea returns dollars to the Fed, the Fed returns the won, and both sides settle the agreed interest. The Bank of Korea keeps whatever spread it charged its domestic banks above what it paid the Fed.
The failure points are not subtle once you look at them directly. If the Bank of Korea has already drawn fifty-five billion dollars and the stress intensifies, five billion remains available. Its domestic banks may need twenty billion more. The gap does not get filled. The central bank must then decide which institutions receive the remaining dollars and which scramble in the open market at punishing rates, a position rather like a fire brigade that runs out of water with two buildings still burning. That allocation decision, made under pressure and with inadequate options, is where the internal structure of the swap line directly determines which private-sector entities lose access first. The central bank becomes a rationing agent, not a lender of last resort in any meaningful sense.
What People Overlook About Activation
There is a widespread assumption that swap lines are essentially automatic: stress appears, central bank draws, problem solved. The reality involves considerably more friction, and the friction compounds at precisely the worst moments.
First, the Fed must agree to activate or reactivate a line. The standing C6 lines are always open, but the capped lines to other central banks have historically been temporary, requiring renewal. A central bank whose line has lapsed, or whose renewal request lands during a period of Fed reluctance, is not drawing anything. When a line's expiry coincides with a period of domestic political resistance to extending Fed credit abroad, as occurred in various forms during Congressional debates over the Fed's emergency authorities, the soft gatekeeping layer that sits above the hard cap becomes at least as consequential as the cap itself, and considerably less predictable.
Second, there is a reputational and signaling dimension that policymakers tend to underweight until it becomes urgent. A central bank that activates its swap line broadcasts, to anyone watching settlement flows, that its banking system has a dollar problem. Some central banks have historically been reluctant to draw for exactly this reason, preferring to let foreign-exchange reserves drain before touching the swap line. A bank that waits too long and then draws near its cap in a single large tranche has effectively spent most of its insurance in one move, and the signal that draw sends can accelerate the very panic it was meant to contain.
Third, the collateral that domestic commercial banks must post to receive dollars from their central bank is governed by local rules, not Fed rules. If a central bank's repo framework accepts lower-quality assets, it may find itself holding collateral that deteriorates quickly during a crisis, creating a secondary exposure entirely separate from the Fed swap itself. The Fed's balance sheet stays clean. The central bank's does not.
The Tier System in Practice
Consider two hypothetical central banks: one in a small export-oriented economy with a thirty-billion-dollar cap, one in a large advanced economy with no cap at all. Both face the same global shock, a sharp sudden rise in dollar funding costs driven by a reversal of carry trades.
The uncapped bank opens its swap line, provides dollars to its banking sector, and the stress subsides domestically. Its banks never see the worst of it. The capped bank draws its thirty billion. Its banking sector needs forty-five billion. The remaining fifteen billion worth of need hits the open market. Dollar borrowing costs for its banks spike. Several smaller institutions face margin calls on dollar-denominated positions. One mid-tier bank requires emergency liquidity support from its government, which in turn draws on foreign-exchange reserves that currency traders had already been watching closely. A run on the exchange rate begins.
The swap line did not fail. It performed exactly as designed. The design, though, encoded a hierarchy, and that hierarchy produced a cascade the capped country's policymakers had not fully modeled. This is not an edge case or a stress-test curiosity: it is the predictable consequence of a system whose architecture most of its beneficiaries have never read carefully.
If you work in a jurisdiction with a capped line, ask yourself this: when did anyone in your institution last treat the number on that cap as a binding constraint rather than a reassuring headline figure? It is a policy judgment about how much stress your banking system is expected to absorb before the Fed's balance sheet becomes your backstop. The judgment was made in Washington, not in your capital.
The Quiet Architecture of Dollar Power
Swap lines are sometimes described as a public good, a piece of international financial infrastructure that stabilizes the global dollar system for everyone's benefit. That framing is not wrong. It is incomplete in a way that matters. The uncapped lines flow to the central banks whose financial systems are most tightly integrated with dollar markets and, not coincidentally, to the economies with the deepest strategic relationships with the United States. The capped lines extend the system's reach without fully extending its guarantee. The difference between those two things is the difference between membership and access.
The internal structure, the caps, the renewal requirements, the activation thresholds, is where dollar hegemony becomes operational rather than rhetorical. It is not a matter of which country the Fed likes. It is a matter of which country the Fed has legally committed to backstop without limit, versus which country receives a defined ceiling and must manage whatever lies beyond it alone. History suggests that managing what lies beyond the ceiling, during the precise moment when the ceiling is reached, is the hardest thing a central bank can be asked to do.
By then, the queue is already forming, and the phones are already ringing.