The Same Three Exits. The Valve Is Where They Differ.

Every sovereign that has to fund a deficit the market will not buy at its chosen price has the same three exits: raise rates, sell reserves, or print. The bind is now global. So is the choice of who pays for it.

Start with the bind. Fiscal dominance means the government's borrowing need sets the ceiling on what the central bank can do, rather than the other way round. Independence is not abolished; it is repriced. A bank that would tighten to defend its currency has to weigh the cost to its own sovereign's interest bill first.

From that bind there are three exits. Raise rates and defend the currency, at the cost of the carry trade and the pension funds holding it. Sell foreign-exchange reserves to buy the local currency, at the cost of the bond market those reserves sit in. Or print, or swap, and let the currency fall, at the cost of imported inflation. No sovereign escapes the bind. It chooses which exit to take, and the choice is visible in the price that moves.

The claim this piece tests is that the exits are the same everywhere but the valve is not. The adjustment does not disappear. It surfaces in the currency in Japan, in the spread in Europe, in the term premium in the United States. And an asset that hedges one valve does not hedge another.

One bind, three exits

Fiscal dominance has a precise meaning. The central bank's policy is constrained by the government's borrowing need, rather than the other way round. A bank that wants to raise rates must first ask what that does to its own sovereign's interest bill and to the currency the sovereign borrows in.

The bind is arithmetic, not politics. Debt service reached 20.9% of US federal receipts in the second quarter, by our own calculation from the BEA accounts. Interest compounds on the stock, the stock grows with the deficit, and the deficit is not falling. At that point the central bank's reaction function and the sovereign's funding cost are the same object, viewed from two sides.

Three exits follow. Raise rates: defend the currency and the inflation target, and accept that the carry trade unwinds and the funds holding it take the loss. Sell reserves: defend the currency with the foreign assets on the central bank's own balance sheet, and accept that those sales land on the bond market of the country that issued them. Print or swap: refuse to choose between the currency and the bond market, let the currency fall, and accept higher import prices.

The exits are not exclusive. Every sovereign uses some mix. What separates one from another is which exit it can lean on without the cost arriving somewhere it cannot survive. That is why the same bind produces a different price in every market.

Where the cost surfaces

The exits are the same. The valve through which the adjustment escapes is not. On the United States the valve is the term premium. On Japan it is the currency. On the euro area it is the spread. On Britain it is gilt volatility. On China it is suppression, because it has the one thing that relaxes the bind - capital controls - and the one thing that makes it unnecessary - deflation.

Read the four large sovereigns by where their 10yr yields sit. The US 10yr is 5.237%, the 96.7th percentile of its own history. The UK 10yr gilt is about 5.34%, a multi-decade high, per trade data on 2 October. The French 10yr OAT is about 4.90% against a German Bund of about 3.60%, a spread of roughly 150 basis points, a level last seen in 2012, per trade data on 2 October. The Japanese 10yr is about 3.05%, down from a 30yr high of 3.10% the week before. The Chinese 10yr is about 1.68%, the only one of the five that is calm.

The dispersion is the finding. 15yr yields, five different prices, one shared fiscal problem. The market is not pricing a common global risk. It is pricing five different answers to the question of who absorbs the adjustment. The United States pays in term premium, because it can. China pays in growth, because its problem runs the other way. Japan pays in yen. France pays in spread. Each is the same bind, priced in a different currency.

Japan: the exit is the currency

Japan is the clean case because the exit is already visible. The Bank of Japan raised its policy rate to 1.25% in September, a 31yr high. Tokyo core inflation then printed at 2.7% in September, from 1.8% in August, per the Statistics Bureau. The measure that strips out energy and fresh food rose to 3.0% from 2.0%. Services inflation, the part the Bank watches for proof the pressure is domestic, ran at 2.3%, the strongest since November 2023.

By the arithmetic of a hiking cycle, that should have lifted the yen. It did the opposite. The dollar was near 158 yen, its third weekly gain. The market read the same data and concluded the Bank of Japan is behind, not ahead: a policy rate of 1.25% against core inflation of 3.0% leaves the real rate deeply negative. A hike that stays behind the curve does not defend the currency. It confirms the currency is the valve.

The cost of that valve is well known. The yen carry trade stays funded, and the Japanese institutions that hold foreign duration stay unhedged abroad rather than repatriating. Japan remains the largest foreign holder of US Treasuries, at $1,103.9bn as of July, per the Treasury's own TIC data. That holding fell by $66.7bn in May, the largest monthly drop in 3 years, and a further $26.4bn in June. The monthly report for August is due 16 October. The point is not that Japan is dumping. It is that Japan, when the yen is its valve, is not the marginal buyer the US long end needs.

Europe: the exit is the spread

The euro area runs the same bind without a currency of its own. Each member state borrows in a currency it does not issue, and there is one policy rate for twenty different fiscal positions. The adjustment cannot escape through the exchange rate. It escapes through the spread between the weakest sovereign's paper and the strongest's.

That spread is moving. The French 10yr over the German Bund reached roughly 150 basis points on 2 October, a level last seen in 2012, the peak of the euro-area sovereign-debt crisis, per trade data. The French 10yr is about 4.90%; the German 10yr about 3.60%. The direction is the same as the US long end's: the price of the sovereign's own paper, not the price of the central bank's policy.

For the ECB the choice is the same three exits, minus one. It cannot sell reserves in a foreign currency to defend its own, because it does not have one. It can raise rates, and risk the periphery. It can buy the spread down, which is quantitative easing by another name, and accept the inflation cost. That cost is what makes the German answer no: a central bank buying the spread down is choosing the inflation exit, and it is the exit the euro area can least afford. A central bank that cannot cap the spread and cannot defend a currency has one valve left, which is the price of the paper - and it prices itself.

America: the exit is everyone else's

The United States has the one exit no one else has. It borrows in the currency the world holds as reserves. So it can run the bind longer, spread it wider, and export the inflation the other exits would force it to absorb at home. The dollar is the valve, and the valve is open.

What the United States cannot export is duration. Foreign buyers can take dollars and leave; the long bond still has to find a marginal buyer at a price. That is why the American valve surfaces as term premium rather than as a currency. Every Treasury coupon from 5yr to 30yr sits between the 95.2nd and the 98.2nd percentile of its own history, as of 1 October. The 30yr is 5.603%, the 97.9th. The 10yr real yield is 2.877% against a breakeven of 2.36%.

The front end tells a different story, and it is the counter-evidence this argument has to carry. The 13-week bill fell to 3.982%, the 25.8th percentile, and secured overnight funding closed to the administered rate at a spread of zero. The bills market accepted the policy path. The selloff is a long-end event, not a funding event. A bind that had reached the funding market would look different: bills backing up through the administered rate, and the reverse-repo facility refilling. Neither is happening.

The honest reading is that the United States is in the bind but not yet at the wall. The adjustment is being paid in the term premium, slowly, and the front end is still functioning. That is the difference between a repricing and a crisis. It is also why the reserve privilege is real and not permanent: the more the United States leans on the dollar to absorb the adjustment, the more of it arrives as imported inflation, and the harder the eventual choice becomes.

Hard assets are the hedge. The regime sets the entry.

The exits explain the demand for assets that are nobody's liability. If the adjustment is paid in depreciating currency, in widening spreads, or in suppressed real yields, then the store of value that cannot be devalued or restructured wins. Gold and bitcoin are the two instruments that sit outside the sovereign's balance sheet entirely. That is the thesis, and it is sound.

It is also incomplete, because hard assets are not one trade. Copper, at the 93.4th percentile of its own history as of 1 October, is a growth asset - a reflation bet, and it fires when the market prices demand, not debasement. Gold is a real-rate asset: it rose $39 on 2 October, to 4,241, but gave back $100.50 over the week in which the 30yr rose 14 basis points, which is the behaviour of a duration-sensitive asset, not a pure debasement hedge. Bitcoin is the high-beta liquidity leg. Silver is half-industrial. They do not fire on the same trigger.

That matters because the regime is not one thing. The composite index this site maintains reads TIGHTENING at 0.350 on 5 of 5 sub-indices, with Rates & Policy dominant, as of 2 October. In a tightening regime, hard assets are early, not wrong. The insurance is cheap precisely because the event has not arrived, and it is cheaper to own the hedge than to time it. The correct frame is not that the portfolio owns too few hard assets. It is that it is short the specific bucket that fires on the specific valve expected to open.

Where this breaks

First, the front end. If US bills back up through the administered rate and the reverse-repo facility refills above $50bn, the bind has reached the funding market and the term-premium read is no longer sufficient. That is the crisis version, and it is not priced.

Second, the currency. If the yen strengthens through intervention rather than through a narrowing rate differential, Japan has chosen the reserve-sale exit, and the US long end loses its marginal-buyer assumption directly.

Third, the spread. If the OAT-Bund spread compresses without the ECB buying, the euro-area bind has relaxed on its own and the spread is not a valve.

Fourth, the privilege. If the dollar falls while US long-end yields also fall, the adjustment is being paid somewhere other than American term premium, and the exceptionalism in this piece is wrong.

Fifth, Japan's inflation. If Tokyo core CPI falls back under 2% next month, the pressure that forced the currency valve is gone, and the cleanest case in the argument stops being clean.

The exits are the same everywhere. What the market is pricing, across five curves, is which door each sovereign walks through, and who is standing behind it when it does.