People throw around the word "crisis" every time yields move. It's worth being precise about what that word actually covers, because the different versions call for completely different responses.
Markets reprice all the time as new information arrives. That's their job. A repricing on its own is not news. What matters is the size of the move, the speed of the move, and whether anything underneath the market itself is breaking. Those questions are what separate the ordinary from an actual crisis.
The historically unusual thing on this chart isn't today's yields. It's how cheap long-term money became after 2008. In the US, UK, Australia, and Germany, yields fell well below the ranges that had prevailed through most of the previous two decades. They've since moved substantially back toward those earlier ranges.
Japan is a different story entirely. Its yields fell after its asset bubble burst in the early 1990s, close to two decades before the rest of the world followed it down. Japan has effectively been in therapy since 1990. Its current climb isn't a return to normal after a shared global episode. It's part of an attempted exit from a much longer, home-grown one.
Three things people mean by crisis
Strip away the word and there are really three separate problems people are pointing at, plus one way any of them can spread.
The first is market dysfunction: an orderly repricing turning disorderly. The UK's 2022 mini-budget is the clearest recent example. Rising gilt yields triggered collateral calls on leveraged pension funds running liability-driven investment strategies, forcing those funds to sell gilts into a falling market, which pushed yields higher and forced more selling. The Bank of England stepped in because that feedback loop threatened financial stability. The underlying problem was forced selling and liquidity, not doubt about Britain's ability to repay its debt. That distinction matters, and it resolved within days once the Bank intervened and the forced selling stopped.
The second is funding failure: demand for a government's bonds drying up. Sovereign auctions almost never literally find no buyers, since price adjusts to clear the market. What people usually mean by a weak or failed auction is exceptionally soft demand, a large tail relative to where the market was trading beforehand, weak bid-to-cover, or heavy dealer absorption, and at the extreme, a government unable to issue except at prohibitively high yields. This hasn't happened to a major developed economy in the current cycle.
The third is a debt sustainability crisis: rates and borrowing reinforcing each other. Higher rates push up the interest bill, the bigger bill pushes up borrowing, and bigger borrowing pushes rates higher again, until some outside adjustment breaks the loop, fiscal consolidation, restructuring, inflation, or external support. Greece in 2009 is the clean example: a funding crisis, a currency the government didn't control, and a loop that finally forced restructuring.
Monetary sovereignty changes the form of this constraint. It doesn't abolish it. A government that can always create the currency its debt is denominated in doesn't need to default for want of cash, but accommodating an unsustainable fiscal position can still push the adjustment elsewhere, into inflation, the exchange rate, or nominal bond yields. It's a different failure mode from Greece's, just not necessarily a milder one.
Contagion isn't a fourth problem sitting next to these three. It's how any of them travels: stress in one country's bonds spreading to others, either because investors treat the risks as correlated, or because the same leveraged players get forced to sell several markets at once to raise cash.
Where we actually are
None of the three problems above fits what's happening now. There's no funding failure. There's no forced-seller spiral like the UK saw in 2022. Nobody has lost market access. What's happening instead is an ordinary, orderly repricing, just a large and sustained one, spread over years rather than days, with no single dramatic session driving it. That's the baseline case against which the three crisis scenarios above are the exceptions, and it calls for governments to accept a higher cost of capital and adjust their budgets to fit, on a timeline measured in years, not an emergency response measured in days.
Don't mistake "not a crisis" for "nothing is happening." Over the past three years the world has swallowed a substantial repricing, and that repricing has forced substantial adjustments.
France is the sharpest current example. German Bunds set the risk-free rate for the whole eurozone, and Bunds have moved briskly of late, up to their highest level since 2011. That lifts the floor under every European government's borrowing cost. France sits well above that floor, on top of its own political troubles. Successive governments have tried to pass deficit cuts and kept failing to get them through a fragmented parliament, so some of that political and fiscal risk is now being priced directly into the spread over Bunds.
The US shows the same pattern without the parliamentary drama. Federal debt held by the public has gone from around 35% of GDP in 2007 to over 100% now, at a similar 10-year yield, and the interest bill is climbing as cheap debt rolls over at today's rates. No politically viable programme has emerged to stabilise the debt path. The Congressional Budget Office's long-term projections are built on current law, and current law doesn't stabilise it.
The common thread is not confusion or denial. It's that adjustment is painful, and the pain is concentrated on identifiable people, while the cost of delay is diffuse and lands later, as a bond market problem rather than a line item anyone voted on. Concentrated pain beats diffuse pain in almost every political system, right up until the diffuse cost gets big enough to force the issue. That is roughly where France and the US both are now.
Why the repricing is happening
For a decade governments borrowed into a world of abundant global savings and enormous central bank balance sheets, where the marginal price of long money kept falling. That world has reversed. A few forces are pushing in the same direction at once.
Private investment is competing harder for capital, with the AI investment boom adding a large new source of demand alongside existing pressures from defence spending, the energy transition, grid investment, and heavy government borrowing itself.
Central banks are no longer absorbing duration through quantitative easing, so more of that new bond supply has to find a home with ordinary investors rather than a central bank balance sheet.
Inflation has come back, which lifts the short end of the curve directly through policy rates, and long yields can rise on top of that through some mix of higher expected future short rates, higher real term premia, and higher inflation-risk premia, alongside genuine uncertainty over how firmly central banks, the Fed included, will hold the line if growth slows.
None of this requires a buyers' strike or a loss of confidence. The market clears in the ordinary way, at a higher yield. A bond market doesn't have to break to discipline a government. The price can simply keep moving until somebody changes behaviour.
The distinction between a repricing and a crisis is the whole ballgame. Confusing one for the other leads to the wrong response, and often to the wrong amount of alarm. But the repricing itself is not nothing. It's the mechanism forcing adjustments that had been avoidable for over a decade, and are not avoidable anymore.
Where does that leave the endpoint? Yields are heading back to normal, in the sense of returning toward the ranges that prevailed for most of the pre-2008 era (Japan and a couple of others excepted). But normal this time won't mean the same numbers as last time. Debt loads are far larger now than they were then, and that larger debt stock is likely to require a higher yield to persuade investors to hold it. The destination isn't necessarily the old range. It may be the old range plus a fiscal premium that wasn't needed last time around.
No comments:
Post a Comment