Mechanism

Duration

Duration measures how much a bond or a portfolio of bonds falls in price when interest rates rise. It is quoted in years, which is the most misleading convention in fixed income, because what it behaves like is a multiplier: a duration of ten means roughly a ten per cent price fall for every percentage point that yields go up. It is the number that converts a rate move into a solvency problem, which is why institutions holding long assets against short money keep dying of it.

How it works

A bond is a set of promised cash flows on future dates, and its price is those cash flows discounted back to today at prevailing yields. Cash flows arriving further out get discounted harder, so the longer the money is out, the more violently the price reacts when the discount rate moves. Duration is the summary statistic for that. It is the weighted average time until you get your money back (Macaulay duration), which conveniently turns out to be almost the same number as the price sensitivity per percentage point of yield (modified duration). Hence the confusion. It is expressed in years and it behaves as leverage.

An illustrative example, with round invented numbers. Take a thirty-year bond with a 4% coupon. Its duration is around 17. If yields go from 4% to 5%, the price falls roughly 16 or 17 per cent — on a $100 bond, about $16 gone, without anyone defaulting on anything. A two-year note in the same move loses about 2%. Same issuer, same credit quality, same one-point rate move, eight times the damage. That is duration.

Two refinements matter. The relationship is not quite linear: the price gain from falling yields is a little larger than the loss from rising ones, a curvature called convexity. And, much more important in practice, liabilities have duration too.

The mismatch is the part that kills you

A life insurer owing pensions in 2055 has enormously long liabilities. A bank funded by deposits that can leave this afternoon has liabilities of approximately zero duration. What kills institutions is not owning long assets; it is owning long assets against short liabilities. That gap is the duration mismatch, and it is the whole of the story.

Why it matters to this crash

The central fact of 2026 is that the price of duration has been repriced upward almost everywhere at once, in an orderly way, without a panic to blame it on. The 30-year Treasury touched 5.33%, the highest since 2007. The New York Fed's sits around 80 basis points, near a twelve-year high, while the of rate volatility was at 9.83 and falling. We wrote that up on 20 August: investors are not frightened of the next Fed meeting, they are declining to lend for thirty years at the old price. That is structural, and policy cannot easily reverse it.

Everyone who already owns duration eats that repricing. Japan's major life insurers, historically the world's most reliable price-insensitive buyers of very long bonds, were carrying ¥30.86tn ($194bn) of unrealised losses on domestic bonds at the end of June, up 60% year on year, with 30-year JGB yields in the 3.9% range. Nippon Life booked ¥44bn of impairments in April–June, Meiji Yasuda ¥25.3bn. We covered it on 19 August. Held to maturity those losses vanish; the risk is that impairment rules, a duration mismatch, or surrendering policyholders force sales first.

Duration is also being manufactured and shuffled around at speed. The Treasury doubled its long-end buybacks from $2bn to at least $4bn per operation, funding them with bills, which takes duration out of private hands today and moves the risk to the refinancing (19 August). The exists because asset managers want duration exposure without spending cash. And the is issuing enormous quantities of it: the Meta Hyperion vehicle's roughly $27bn of debt at 6.58% amortises to 2049, against GPUs most operators depreciate over about five years (20 August). That is duration in the other sense, where the asset's useful life is a fraction of the loan's, and it is being sold to buyers who want long investment-grade cashflow.

What would make this dangerous

Japanese life insurers turning from buyers into net sellers of long JGBs, or impairment charges materially exceeding the ¥44bn and ¥25.3bn already booked, would remove the anchor bid from the global long end. A rising surrender rate is the early tell, because surrenders force sales regardless of what the accounting says. Sony Life's went up 0.2 points to 1.4% in April–June.

Rate volatility rising alongside yields would be the second signal. The current combination of multi-decade-high yields and a very low MOVE says the is repricing calmly. If MOVE climbs off 9.83 while auctions keep tailing, the repricing has become a scramble, and leveraged duration positions get margined.

And watch funding. at 3.65% has not moved. If it starts printing above 's floor, is tightening, and the largest holder of leveraged duration in the market becomes a forced seller of it.

As seen in

Every dispatch we have filed that touches this. Newest first.

The Kauri market reopens after nine years2026-08-26
AI capex has stopped being an equity story and become a competitor to governments for the world's fixed-income capacity.
A put you can see in the swap spread2026-08-25
A backstop in the long end suppresses the one price that was disciplining the fiscal and AI-borrowing complex, and moves the adjustment into the currency instead.
"That isn't a crisis. It is an invoice."2026-08-25
The Treasury is now defending a price in the largest bond market in the world, and the critics include the man who taught the Treasury Secretary how to attack one.
The long end repriced, then stopped panicking2026-08-25
A market that needs official buying to clear, and whose marginal buyer is levered overnight, is calm in a way that can end in one session.
A twist of $4bn against a debt of $40tn2026-08-24
The Treasury can change who holds the duration, but not how much borrowing there is, and the market spent three sessions demonstrating the difference.
Citadel names it: the pressure moves, it does not vanish2026-08-24
A calm long end achieved by intervention is not the same as a calm long end, and the difference is showing up in gold and the dollar.
Japanese households take the OpenAI duration risk2026-08-24
When bank credit committees decline AI-related duration and retail investors take it instead, the loss-absorbing layer of the boom moves to households who cannot mark it or sell it.
Paying for the buybacks out of the cash tin2026-08-24
If the government's own liquidity buffer becomes a market-management instrument, the shock absorber and the shock are drawn from the same pot.
The pressure has to come out somewhere2026-08-24
When a government suppresses the price of its own debt, the adjustment does not disappear — it relocates to the currency, and gold is where you watch it happen.
A 19-year high, reached without a panic2026-08-23
The long end has settled at a level that reprices every leveraged balance sheet in the system, and it got there calmly enough that nobody has been forced to do anything about it yet.
Four billion a go, against two trillion of supply2026-08-23
The Treasury has now revealed the yield level at which it feels compelled to act, and the market erased the effect within 24 hours.
The Treasury and the hyperscalers are in the same queue2026-08-22
AI capex and the federal deficit are now competing for the same duration buyers, which makes the long end of the Treasury curve the binding constraint on the buildout.
Forty trillion, two years ahead of schedule2026-08-21
A Treasury that intervenes to cap its own borrowing costs converts a bond-market problem into a currency problem, which is roughly what gold is pricing.
A data centre bond that amortises to 20492026-08-20
The AI buildout's debt is being termed out to 2049 against assets that turn over every few years, and it is being placed with buyers who do not mark it daily.
Everything that isn't a dollar went up2026-08-20
A Treasury that manages the yield curve directly transfers the adjustment to the currency, and the gold price is where you see it first.
Forty trillion, and a buyback JPMorgan doesn't believe2026-08-20
The government is managing its borrowing cost by shortening the maturity of its debt, which lowers today's yield and moves the risk to the refinancing.
Japanese savers are quietly leaving the yen2026-08-20
The marginal buyer of long-duration paper for three decades was a Japanese institution with nothing better to do with the money; it now has something better to do with the money.
Nobody fears rates. They want paying to hold them.2026-08-20
A high term premium with low rate volatility means the long end is being repriced structurally, not panicking — and buybacks address panic, not structure.
A bank designed so the deposits can leave at 3am2026-08-19
The 2023 bank runs were the fastest in history because of technology; this one is being built with the friction deliberately removed.
One hand buys bonds, the other stops2026-08-19
The government is managing the long end by shifting its borrowing to the front end at exactly the moment the Fed has stopped supplying reserves there.
Personal injury claims are now a bond2026-08-19
When the ABS market starts financing cash flows this exotic, it is a statement about how much money is looking for yield rather than about how good the collateral is.
$194bn of losses Japan's insurers don't have to show
The largest patient buyers of long-dated government bonds are sitting on losses that only stay invisible if they never have to sell.
$830bn of borrowed money sits in the bond market
The largest single leveraged position in global finance is a bet on Treasury market functioning, at the moment Treasury market functioning is the question.
Every long end at once
Term premium is rising everywhere at once, and the marginal holder of US duration is a repo-financed relative-value trade rather than a real-money investor.
Japan's insurers are sitting on $194bn
If Japanese life insurers become sellers of duration rather than buyers, the global long end loses its most reliable price-insensitive bid.
Japan's insurers are sitting on $194bn of losses
The marginal global buyer of very long bonds is becoming a manager of its own losses, which removes demand precisely where supply is exploding.
Japan's insurers hold $194bn of paper losses
Japan's life insurers are the marginal global buyer of very long bonds; if rates force them to shorten duration, the long end everywhere loses its anchor.
Japan's life insurers are $194bn underwater
The marginal buyer of long-dated global debt is sitting on losses that could force it to become a seller instead.
Multi-decade high yields, and the calmest bond market in a month
The long end is absorbing government and AI borrowing simultaneously, and the price of that absorption is a permanently higher discount rate for everything else.
Nobody in the survey mentioned the basis trade
The Treasury market's biggest hidden buyer is leveraged and funded overnight, which makes a funding squeeze, not a fiscal event, the likelier trigger for disorder.

Written 2026-08-20. Crashopedia entries are drafted from sourced evidence, fact-checked against it, and edited by hand. If something here is wrong, it is wrong in git and can be fixed there.