The 30-year US Treasury bond is not behaving.
On 19 May its yield touched 5.198% — the highest since July 2007, which is to say the highest in nineteen years. It has eased back a little since. The exact level is not the point.
The point is what the long bond is. It is the anchor — the price of money over the long term, the nearest thing markets have to a risk-free rate, the benchmark every other asset is priced against. A reference point, not a point of interest. For decades, that was exactly what it was. Lately it has behaved less like an anchor than like a risk asset: volatile, selling off, repricing at speed.
That is the misbehaviour of the title. And in this market, misbehaviour is worth a closer look — because of what the long bond touches, and because of what tends to be reached for when it will not settle down.
The market that takes no instruction
Central banks set interest rates — or so the popular description goes. At the short end it is roughly true: the central bank steers the overnight rate through open market operations, buying and selling short-dated government debt, and the inference has always been that the rest of the curve follows the lead. Usually, it does.
The long end is another matter. No one sets the 30-year yield; it is subject to market forces. A central bank can influence it — and for fifteen years they worked hard at exactly that, through bond purchases, forward guidance, and the steady signal that the official hand was on the tiller. For most of that period, the influence held. The curve was compliant. Investors in the long-term obligations of Western governments were grateful for the income and the assurances that came with it.
The long bond — or rather the investors who hold it — now has an opinion of its own, and it diverges from the central bank's. Inflation has not behaved as forecast. Deficits grow regardless of who is in office. The supply of new debt has to find a willing buyer somewhere. At the long end, a repricing is under way: long-dated debt is not as attractive as it recently was. It is a view unsanctioned by any central bank or political party.
This matters more than the average price change, because of what the long bond sits beneath. It is the reference rate for mortgages, for corporate borrowing, for the cost of government debt itself. It is the discount rate under the valuation of almost every other asset. If a long-dated government bond pays a 'risk-free' 5%, that is the hurdle rate every risk asset now has to clear.
It is also, quietly, a destabilising force when it rises. There is a great deal of leverage in the financial system, most of it arranged at lower rates on the implicit expectation that rates would stay low — or fall further. When the anchor rate climbs, the structures built on the assumption it would stay put come under strain. Rising long rates have a long history as the catalyst — not always the cause, but the catalyst — of trouble elsewhere.
The cleanest recent example is the autumn of 2022, when UK pension funds running liability-driven strategies were leveraged to low gilt yields. Long yields jumped; the leverage turned against them; forced selling drove yields higher still, which forced more selling. The Bank of England had to step in within days. Nobody had it on their list of likely crises. That is rather the point — the long bond is where this kind of thing starts.
A policy of control without controls
Suppose the misbehaviour continues. Suppose the long bond keeps repricing, keeps climbing, keeps unsettling the structures around it. What does a central bank do?
It has a tool for this. Yield curve control.
Quantitative easing — once novel, now routine — buys bonds by the quantity: a central bank announces it will purchase, say, $80bn a month, and it does. Yield curve control dispenses with the quantity. It names a yield instead. The central bank picks a level — 4%, 3%, whatever it decides — declares that the relevant bond will trade there, and commits to buy however much it takes to make that true. However much. There is no announced ceiling, because the absence of a ceiling is the point: an unlimited bid, standing in the market, daring anyone to sell.
Price fixing has always been an attempt to reconcile two variables: you can control the price of something, or its supply, but never both at once. The supply of government debt is guaranteed — and growing. Under yield curve control the central bank claims it can guarantee the price as well.
This ups the ante. Inflation rises, real yields fall, and fixed income — traditionally a portfolio's conservative ballast — shifts firmly into the risk asset category. Whether its largely passive holders go along for the ride is an open question. There is also the deeper point that price-fixing always produces a fault line. If both the price and supply of government debt are guaranteed, the pressure has to go somewhere. It tends to land in the currency.
No one is proposing this for the United States today, and that matters — it is a tool in the box, not a policy on the table. But it is a real tool, not a hypothetical one. The Federal Reserve discussed it openly in 2020. And the question of what sits in the box becomes a good deal more interesting when the long bond behaves the way it has lately. It is worth knowing what the tool actually does. The best way to know that is to look at the three times it has been used.
No clean exit
Yield curve control is not new. It has been used three times by credible central banks. The episodes are worth knowing in some detail, because together they tell you what the policy's advocates tend to leave out.
The United States, 1942 to 1951. To finance the Second World War, the Federal Reserve pegged the entire curve — around three-eighths of one percent on Treasury bills, a cap of two and a half percent on long bonds — and bought whatever was needed to hold those levels. As wartime policy it did the job. The difficulty came afterwards. The war ended in 1945; the peg did not. It ran for six more years, because once the government had grown used to financing itself at a guaranteed low rate, the Fed found it could not simply stop. Ending it took a direct institutional confrontation between the Federal Reserve, which wanted its independence back, and the Truman administration, which wanted the cheap money kept. The fight was settled by the Treasury–Federal Reserve Accord of 1951 — the agreement now generally regarded as the founding document of the modern, independent Fed.
Japan, 2016 to 2024. The Bank of Japan targeted the ten-year government bond yield at approximately zero. It set a band around the target; then, as the market leaned on the band, it widened it; then widened it again. The policy ran for eight years. It was formally ended in March 2024 — and even then, the Bank kept buying government bonds by the trillion of yen each month. Japan's lesson is not drama but duration: yield curve control became part of the furniture, and the exit took the better part of a decade and was never quite complete.
Australia, 2020 to 2021. The Reserve Bank of Australia capped the three-year yield at one-tenth of one percent. This is the short episode, and the instructive one. In October 2021 the market simply decided to test it. The targeted bond's yield broke above the cap and kept going — the largest monthly move in nearly three decades. The Reserve Bank defended the line once, with an unscheduled purchase, and then — faced with the cost of defending it properly — stopped, and abandoned the policy within weeks. A later review was candid about the damage done to the institution's credibility.
Three central banks. Three different decades, three different reasons, three different designs. And one feature common to all of them: not one exited cleanly. The United States needed six years and a constitutional-scale fight. Japan needed the better part of ten years. Australia lost control of the policy and walked away from the wreckage.
Yield curve control is easy to begin. That is its appeal — it works, at first, and it works dramatically. What none of the three precedents offers is an example of it being ended on the central bank's own terms, at a time of its own choosing, without cost.
What it would mean
So suppose it happens. What would it mean?
The first consequence is inflation, and the mechanism is not subtle. A central bank that caps the nominal yield on government debt while prices in the economy are rising has made a comprehensive decision: it will no longer let the cost of money rise to meet inflation. It will buy whatever it takes to hold the yield down. That is not a side effect of yield curve control. It is what yield curve control is — debt monetisation, conducted openly, under a technical name. Where the inflation then lands — consumer prices, asset prices — can vary. That it lands somewhere does not. American inflation ran high through the second half of the 1940s, with the peg in place. That was the mechanism on display, not a coincidence beside it.
The second consequence is for real yields. The price of gold responds, among other things, to real yields — the return on a government bond after inflation. High real yields make gold, which pays nothing, costly to hold. Low or negative real yields remove that objection.
Yield curve control has a direct consequence here. The nominal yield is held at target through open market buying — the central bank commits to being the marginal buyer at the chosen level, absorbing whatever the market offers at whatever scale that requires. What it does not commit to is controlling inflation. A policy that holds the nominal rate down while prices rise produces negative real yields. Not as a side effect — as arithmetic.
Gold has spent much of 2026 correcting from its January high; with inflation fears pushing real yields up, that correction is rational. Markets are pricing current conditions. That is what markets do. The structural point runs the other way: a central bank acting as marginal buyer in an inflationary environment compresses real yields by design — and removes the primary objection to holding hard assets.
The third consequence is the hardest to measure and the most important. The long bond is not only a borrowing rate; it is a price — and what it ultimately prices is confidence in the long-run management of money. Borrow sensibly and guard the currency, and the market rewards you with a low, steady long yield. Do the opposite, and it charges you a rising one — and, in time, a more volatile one.
Yield curve control does not change that. It changes the number. And a number held down by official buying tends to need more official buying to stay there. Intervention has a way of becoming the condition it was meant to cure: each round makes the next dose larger, the dependence deeper, the exit narrower. That is not a feature of this policy in particular. It is a feature of the instinct behind it.
The instinct to manage
A 30-year yield at a nineteen-year high is the market saying something.
The instinct of policy, when a market says something inconvenient for long enough, is to manage the market rather than hear the message. It is the more comfortable option. It postpones the reckoning. It can be explained in technical language that sounds like competence. And interventionist monetary policy is hardly a radical idea any more. It works, too — at first, and dramatically — which is exactly what makes it tempting and exactly what makes it dangerous.
Yield curve control is what that instinct looks like when it is given a name and a mandate. Whether the Federal Reserve reaches for it is not something I am going to predict. The pressure to act on an inconvenient price is real, and in a leveraged system with rising rates it does not go away on its own. But the long bond may also settle by itself; it has done so before.
Jim Grant, who publishes Grant's Interest Rate Observer, has called what replaced the gold standard the PhD standard of monetary policy. The phrase fits. Yield curve control would be a particularly bold instalment of it — improvisation under a technical name.
Fixed income is supposed to be dull. Bond investors are not the excitable type. When they start voting with their feet, it is worth paying attention.