Theme · past 12 monthsUS, UK, Euro area, Japan

Term premium returns: long yields at multi-decade highs

Hikes explain the front end; long yields at multi-decade highs are a term-premium story of inflation risk, supply and fiscal worries.

What it did to the curve

Front end

Yields higher

2-year

Long end

Yields higher

10–30-year

Curve

mixed

long minus short

Long ends hit their highest in decades in the US, UK and Japan (since 2011 for Bunds), but curves split: US 2s10s flattened (about +52bp to +37bp) and 10s30s flattened too (about +58bp to +35bp) as markets priced the Fed’s turn to hikes, while JGB 2s10s steepened (about +78bp to +110bp).

US 10-year, 7 Oct peak

~5.35%

highest since 2002

Up from 4.05% a year earlier. It had eased to 5.085% by 16 Oct, still about 104bp higher on the year.

UK 30-year gilt, 7 Oct peak

6.036%

highest since 1998

It first broke 6% on 1 Oct, at 6.029%. The 10-year gilt first went above 5% in March.

Bund 10-year, 16 Oct

3.42%

+84bp on the year

It first went above 3% in March, the highest since 2011: even the euro area’s benchmark long bond repriced.

ACM 10-year term premium

~1.07%

9 Oct, model estimate

The extra yield investors demand to hold a 10-year rather than roll short bills. That day the 10-year closed at 5.24%.

What happened

Hikes explain the front end, not 30-year gilts at 6%. A long yield is the expected path of short rates plus a : the extra return investors demand for locking money up when inflation, supply and politics are uncertain. This year all three pushed the same way.

The energy shock raised the risk that inflation stays high. Deficits stayed large, with the CBO putting the US FY2026 deficit at about $2.0tn. And central banks bought less: the BoJ kept and the BoE sold long gilts until September. That leaves more for private investors, who want paying for it.

The result was the highest long yields in decades in the US, UK and Japan, and since 2011 in Germany. The US 10-year reached about 5.35% on 7 October, the highest since 2002, and that day’s 10-year auction priced at 5.300%. The 30-year went above 6% for the first time since 1998, peaking at 6.036%. The Bund 10-year broke 3% in March, and Japan’s 30-year set a record 4.235%.

Curves didn’t simply steepen, which is the trap. The turn in expected policy lifted the 2-year about 119bp. It pulled the 10-year up about 104bp, but the 30-year, furthest from policy, only about 80bp. So US 10s30s flattened from about +58bp to +35bp, while Japan’s 2s10s steepened.

Cross-market spreads show term premium more cleanly. The US 10-year rose more than the Bund even though the ECB hiked more, so policy can’t explain the gap.

Mid-October showed the counter-force. With yields this high, money rotated out of tech stocks into bonds, and the US 10-year eased to 5.085% by 16 October. At these levels, buyers appear.

Timeline

14 Apr 2026 · UK
Record £15bn 10-year gilt syndication at 4.9158%, the highest yield on a 10-year gilt sale since 2008
2 Sep 2026 · Euro area
Global bond sell-off: Bund 10-year at about 3.4%, the highest since 2011
17 Sep 2026 · UK
BoE halts long-dated gilt sales; gilt yields fall about 10bp
24 Sep 2026 · Japan
JGB 10-year hits 3.055%, the highest since August 1996
1 Oct 2026 · UK
Gilt 30-year tops 6% (6.029%) for the first time since 1998
7 Oct 2026 · US
US 10-year reaches about 5.35%, the highest since 2002; the 10-year auction prices at 5.300%
15 Oct 2026 · US
Treasuries rally as money rotates out of tech stocks into bonds

What could change it

Base case

Long yields stay high but stop making new highs. Hikes are mostly priced, and buyers keep showing up at these levels, as the 7 Oct 10-year auction (bid-to-cover 2.77) and the mid-October rotation into bonds showed.

Watch: FOMC, 27–28 Oct · BoJ, 29–30 Oct

Term premium rebuilds

A Budget that fails to reassure on headroom, oil pushing higher and hot US inflation send long ends back through their October highs, led by 30-year gilts and French OATs.

Watch: Moody’s on France, 23 Oct · UK Budget, 28 Oct · US PCE, 29 Oct

Relief rally

A Hormuz deal pulls oil down and inflation fears fade, while the rotation from stocks into bonds continues. Long ends rally hardest and curves flatten from the long end, as in the week of 12–16 Oct.

Watch: UK CPI, 21 Oct · UK Budget, 28 Oct · FOMC, 9 Dec

How desks express it

US 5s30s steepener

Curve

Receive 5y, pay 30y SOFR swaps (DV01-neutral)

If deficits and supply keep rebuilding term premium, the 30-year should cheapen against the 5-year.

1
Term premium sits mostly in long maturities, so supply and fiscal worries hit the 30-year harder than the 5-year.
2
Supply isn’t shrinking: the CBO puts the FY2026 deficit at about $2.0tn, $218bn more than FY2025.
3
Receiving 5y against paying 30y, DV01-neutral, makes it a that isolates that premium from the level of rates.

Wrong if: Hikes lift the belly: the 5-year carries more of the Fed path, so a hawkish FOMC on 27–28 Oct or 9 Dec would flatten 5s30s. More rotation from stocks into bonds, as in the week of 12–16 Oct when the 30-year rallied most, would also richen the long end.

UST–Bund 10-year widener

Cross-market

Sell 10y Treasuries, buy 10y Bunds (DV01-neutral)

The US 10-year has cheapened against Bunds even though policy pointed the other way, which looks like US term premium.

1
Policy favoured Treasuries: the ECB added 50bp over the year while the Fed’s upper bound ended 25bp lower.
2
Yet the US 10-year rose about 104bp against 84bp for the Bund, leaving the spread at about +167bp on 16 Oct.
3
Heavier US supply, with a deficit of about $2.0tn, is one explanation; if that premium keeps building, the spread widens.

Wrong if: Foreign demand for Treasuries is strong (indirects took about 80.3% of the 7 Oct 10-year auction), and a Fed that signals it is done on 9 Dec would rally Treasuries. A hawkish ECB on 29 Oct would lift Bund yields and narrow the spread.

Educational expressions of the theme — how a desk might position, not advice.

Test yourself

Finishing the quiz adds this theme’s key facts to your review deck.

The stories behind it

Sources