2Y 4.800 ▼5Y 4.900 ▼10Y 5.100 ▼30Y 5.460 =VIX 25.46 ▲MOVE 101.88 ▲DXY 103.60 ▲USD/JPY 159.49 ▼GOLD 3,852 ▼COPPER 6.01 ▲WTI 85.84 ▲BTC 64,856 ▼ETH 1,819 ▼S&P 7,341 ▼NDX 27,002 ▼MSTR 96.50 ▼
Rates · Liquidity · Scarcity · Credit · Power · 2026-11-17

The Long End Got Cheaper While 12 Months Got Dearer. That Split Is the Political Map.

A government bond carries 2 prices inside one yield. One is the average of the short rates the buyer expects over the bond's life. The other is the term premium: the extra compensation for locking money up while the future is unknown, which is where supply, inflation risk and political risk all get charged. Split a yield into those 2 parts and a political event stops being a direction and becomes a location. A result that changes how much the state must borrow lands in the term premium. A result that changes who sets the policy rate lands in the expected path, and the expected path is priced at the front, not at the 30yr. The 3 November midterms did one of those and not the other, and the tape has now had 7 sessions to say which. The answer is unusually clean: the duration premium fell 21.1 basis points in a month while the 1-year-ahead forward rose 38.0 in 4 sessions. This piece prices that split, names the 2 buyers on either side of it, and states the level at which the reading breaks.

The split is 38 basis points at 12 months and 17 at 5 years

Start with the forwards, because a forward rate is the only clean way to ask what the market thinks about a specific horizon. Between the 6 November and 12 November par closes, the 1-year-ahead 1-year rate - the rate implied for the period from November 2027 to November 2028 - went from 4.69% to 5.07%, plus 38 basis points. The 2-year-ahead 1-year rate rose 23 to 4.98%. The 3-year-ahead 2-year rate rose 31 to 5.03%. Then the decay starts: the 5-year-ahead 5-year rate rose 17 to 5.25%, and the 10-year-ahead 10-year rate rose 14 to 5.86%.1

change in the forward rate, bp, 6 Nov to 12 Nov closeamber is the 1-year-ahead forward | the move decays as the horizon lengthens010203040381y1y5.07%232y1y4.98%313y2y5.03%175y5y5.25%1410y10y5.86%level under each bar is the 12 Nov forward | source: Treasury par curve, bootstrapped4-session change in forwards. A forward is a rate implied by 2 par yields; it is not a forecast.

The shape is monotonic from 12 months outward and that is the whole argument. A repricing driven by duration supply does not look like this. It concentrates at the back, because the back is where the coupon calendar lives, and it drags the long forwards up faster than the near ones. What happened here is the reverse: the near forward took 38 basis points and the 10-year-ahead took 14, so 2.7 basis points of near-horizon repricing for every 1 at the far end. The horizon that moved is the horizon over which a policy committee can act and be reversed.

Measured from the 5 November close, before the week's data, the 1-year-ahead forward is up 43 basis points and the 5-year-ahead 16. The extra 5 basis points on the near leg, taken between 5 and 6 November, is the session the front end moved first and hardest.

The 1yr sits above 198 closes and below 19

5.07% is a high number for the 1-year-ahead forward and an ordinary one for 2026. The series has printed between 3.28% on 27 February and 5.251% on 28 September across 217 sessions, with a median of 4.30%. Today's print is the 92th percentile: 19 sessions sit at or above it, and 198 below.1 So the level is near the top of the year but not at it - the September cluster was higher, and it was higher while the 10yr was also higher.

1y1y minus 5y5y, bp, 217 sessionsmedian -59 (dashed) | 8 sessions all sit in September, then 12 Nov at -18-120-90-60-300-18 on 12 Novblue: the 8 September closes at or above -20JanMarJunSepNovGap between the 1-year-ahead and 5-year-ahead forward rates. A gap near 0 means the near horizon prices above the far one.

The gap is the better instrument, because it nets out the level of the whole complex. The 1-year-ahead forward minus the 5-year-ahead forward printed -18 basis points on 12 November, against -45 on 5 November and a -59 median across the file. Only 8 other sessions sit at or above -18, and every one of them falls between 16 and 28 September, when the same 2 forwards were converging for a different reason. The 2026 range runs from -118 on 2 January to -2 on 21 September.1

Read that number plainly. A negative gap means the market prices the year starting 12 months out below the 5 years starting 60 months out, which is the normal shape when the policy rate is expected to fall over time. A gap of minus 18 rather than minus 59 says the market has given up most of that expectation: it now expects the next 2 years to look roughly like the next 10, rather than to start high and fall. That is not a growth forecast. It is a statement that the committee's room to cut has narrowed.

The duration premium fell 21 basis points in the same month

Now the other leg. The ACM 10yr term premium - a model estimate published by the Federal Reserve Bank of New York, which subtracts the expected path of short rates from the 10yr yield and calls the remainder the price of duration risk - printed 0.8738% on 6 November against 1.0844% on 6 October and a 2026 peak of 1.0965% on 5 October.2 That is minus 21.1 basis points in a month in which the governing party changed.

10yr term premium, percent, 214 printspeak 1.0965 on 5 Oct | 0.8738 on 6 Nov | median 0.75780.50.70.91.11.0965 peak0.98570.8738JanMarJunSepNovACM term premium: the compensation for holding a 10yr bond rather than rolling a short bill. Last print 6 Nov.

The level is still high. Against the 214 prints of 2026 the 0.8738% reading is the 78th percentile, 11.6 basis points above the 0.7578% median and 41.1 above the February low of 0.4627%. Twenty sessions this year printed at or above the 16 October level of 0.9857%, all of them in the September-October window.2 So the move is a partial retracement of a 6-week build-up, not a collapse.

Two caveats belong here rather than in a footnote. The series last printed on 6 November, so it is 4 sessions behind the par curve and cannot be used for same-day attribution of the week's 38-basis-point forward move. And the same month's breakeven - the gap between the nominal 10yr and the inflation-protected 10yr, which is the market's price of expected inflation - held at 2.24% on both 5 and 12 November, inside a 2.18% to 2.50% range it has held all year.3 The premium that fell was not an inflation premium.

The mechanism: a closed fiscal channel leaves one variable standing

The 3 November result took the Senate and is projected to take the House. A majority that cannot reach 60 votes in the Senate cannot pass a budget resolution that adds duration supply, and the Treasury's 4 November refunding held the 3yr, 10yr and 30yr offering sizes flat for a fourth consecutive quarter.4 That is the fiscal channel closing in public, in the same week, on the same document. When the supply of duration is fixed, the compensation for holding it falls - which is what the term premium did.

the chain, and the joint where the political risk landsMIDTERMS3 NovdividedgovernmentSENATE60 votesfiscal channelclosedTERM PREMIUM0.8738minus 21.1bpin a monthPOLICY PATH1y1yplus 38bp in4 sessionsBORROWER5.07%1-year fundingrateDuration got cheaper while the next 4 quarters got dearer.1y1y forward4.69 to 5.07+38bp2y1y forward4.75 to 4.98+23bp5y5y forward5.08 to 5.25+17bp10y10y forward5.72 to 5.86+14bp10yr term prem1.084 to 0.874-21bpforwards 6 Nov to 12 Nov | term premium 6 Oct to 6 Nov, the last print | par closesChain is arithmetic, not a forecast. Term premium box is a month, the forward boxes are 4 sessions.

What the result did not change is the policy rate. The Federal Reserve sets it regardless of who holds the House, so the political variable that survives a split Congress is not the deficit. It is who runs the institution and how fast a decision can be reversed. The Fed's own board listing, read on 30 October 2026, names Kevin Warsh as Chairman and lists Jerome Powell among the governors - a structure in which the chair's preferences and the board's composition carry the reaction function.5 A variable of that kind prices at the horizon where appointments, dissents and reversals actually bite: 12 to 24 months, not 30 years.

That is the separation this piece is about. The long end lost its political content because the fiscal channel closed, and the near horizon gained political content because the monetary channel did not. Both moves are political. They run in opposite directions, and they are measured by different instruments: the term premium at the back, the 1-year-ahead forward at the front.

The instrument is the spread, and it is 18 basis points wide

The call is that this repricing is a policy-path event, not a supply event, and the instrument that says so in one number is the 1y1y-minus-5y5y gap at -18 basis points. It is a spread between 2 rates built from the same curve on the same day, so it is immune to the level of the whole complex, to the duration of the week's data calendar, and to the stale term premium series. It moved 27 basis points in 4 sessions, from -45 to -18, while the 10yr par yield rose 24 and the 2yr 35.

forward rates indexed to 1 Sep 2026 = 100amber 1y1y | blue 5y5y1001021041 Sep5 Nov12 Nov+2.9%+0.7%Indexed view, not a forecast: it shows which horizon repriced, not where either rate goes.

The supporting reads are all on the same side. The 1y1y-minus-2y1y hump at 9 basis points is the 47th highest of 217 sessions against a 2 median, so the peak of the forward curve is at 12 months and not at 24 - a curve that expects the first move sooner than the second. The 30yr-minus-10yr spread at 33 basis points is the 19th narrowest of the file against a 53 median, so the back end did not participate.1 Credit did not either: high-yield OAS at 3.30% and investment grade at 0.90% on 11 November were 12 and 3 basis points tighter than on 6 November.6

One number complicates the story and belongs in the body. The 20yr closed at 5.48% against 5.43% at the 30yr, so the 20yr sits 5 basis points over the 30yr - the 11th time in 217 sessions that the gap has reached 5 basis points or more. A market paying more for year 20 than year 30 is not a market that has stopped worrying about duration supply. It has stopped worrying about the last 10 years of it. That is a narrower claim than 'gridlock is priced', and it is the claim the evidence supports.

Where the political risk actually sits on the curve

Put the 2 legs on one arithmetic frame. Over the 4 sessions to 12 November the 1-year-ahead forward rose 38 basis points and the 5-year-ahead 17 - a ratio of 2.2 to 1. Over the month to 6 November the term premium fell 21.1, and the same 10yr par yield that carries it fell -41 basis points from the 6 October close.1,2 Duration got cheaper while the next 4 quarters got dearer, and the 2 facts are not in tension once the channels are separated.

This is the governance read in one line: the election removed a political variable from the long end and left one standing at the front. It is a less comfortable read than 'gridlock is bullish duration', because the same arithmetic produces a higher cost of money for anyone who borrows at 12 months while it produces a lower cost for anyone who lends at 30. A company rolling working capital sees the first. A pension fund buying duration sees the second. Both are looking at the same election.

Where this breaks

The separation read dies if the 1y1y-minus-5y5y gap returns to -45 basis points or below inside 10 sessions. That is 27 basis points of re-widening, it takes the gap back through the whole November range, and it would mean the near horizon was repricing a liquidity or positioning event rather than a policy path. The level is -45 basis points on the 1y1y-minus-5y5y spread, and the trigger is a daily par close at or under it.

The 'duration got cheaper' read dies if the ACM 10yr term premium prints above 0.9857%, the 16 October level, on its next release. That is 11.2 basis points of re-widening from the 6 November print of 0.8738%, and it would say the fall was a 5-session wobble inside a 6-week build rather than a repricing of the fiscal channel.

The policy-path read dies if the 10yr breakeven moves more than 5 basis points from 2.24%. A breakeven above 2.29% would put part of the move back into inflation compensation, and the claim that this is a real-rate event would stop holding. A breakeven below 2.19% would say the market is pricing a growth problem rather than a policy one, which is the same falsification from the other side.

All 3 holding is the confirmation: the gap above -45, the term premium under 0.9857%, the breakeven at 2.24%, and the 1-year-ahead forward at or above 5.07%. Four prices, 2 actors, and the actor at the front is the one that can be replaced by an appointment.

Notes

1. Par yields: data/treasury_par.csv, US Treasury daily par yield curve, 217 sessions from 2 January to 12 November 2026, refetched from the Treasury's yearly CSV on 13 November and merged newest-first; the 12 November row is the newest and was absent from the local ledger before the refetch. All levels are par closes, not intraday marks. Session changes are close to close. 12 November: 3mo 4.32, 1yr 4.65, 2yr 4.86, 3yr 4.90, 5yr 4.95, 7yr 5.02, 10yr 5.10, 20yr 5.48, 30yr 5.43; 2s10s 24.0bp. 6 November: 3mo 4.19, 1yr 4.33, 2yr 4.51, 3yr 4.59, 5yr 4.64, 7yr 4.74, 10yr 4.86, 20yr 5.29, 30yr 5.26; 2s10s 35.0bp. 5 November: 1yr 4.28, 2yr 4.46, 5yr 4.59, 10yr 4.84, 30yr 5.25.

2. Forward rates are computed by bootstrapping the par curve: the b-year rate a years forward solves (1+s(a+b))^(a+b) = (1+s(a))^a x (1+f)^b on par yields rather than zero-coupon yields, which is an approximation that biases each forward by a small amount and does not affect the comparison between 2 dates on the same construction. 6 November: 1y1y 4.690, 2y1y 4.750, 3y2y 4.715, 5y5y 5.080, 10y10y 5.722. 12 November: 1y1y 5.070, 2y1y 4.980, 3y2y 5.025, 5y5y 5.250, 10y10y 5.861. 1y1y series: max 5.251 (28 Sep), min 3.280 (27 Feb), median 4.300, 19 sessions at or above today's print. Gap series: median -59.1, min -118.2 (2 Jan), max -1.9, 8 sessions at or above -18.0, all between 16 and 28 September. Hump (1y1y minus 2y1y): today 9.0, 47 of 217 sessions at or above it, median 2.0. 30yr-minus-10yr: today 33.0, 19 of 217 sessions at or below it, median 53.0. 20yr-minus-30yr: today 5.0, 11 sessions at or above 5.0.

3. Term premium: FRED THREEFYTP10, the ACM 10yr term premium, 0.8738% on 6 November, 1.0844% on 6 October, 0.9857% on 16 October, 1.0965% peak on 2026-10-05, 0.4627% trough on 2026-02-27, median 0.7578 across 214 prints. Percentile 78.0. No print after 6 November, so the series is 4 sessions stale against the par curve and is used for a month-over-month direction, never for same-day attribution. Breakeven: FRED T10YIE, 2.24% on 12 November and 2.24% on 5 November, 2026 range 2.18% to 2.50% across 217 prints.

4. Election and issuance: the result recorded in this file is a Senate flip with the House projected, reported in the 4 November and 5 November pieces. The 60-vote cloture constraint and the flat 3yr, 10yr and 30yr sizes at the 4 November refunding are as stated in the 10 November piece. No claim is made here about appropriations timing.

5. Board composition: federalreserve.gov board members listing, read 30 October 2026, naming Kevin Warsh as Chairman and listing Jerome Powell among the governors. No claim is made here about voting behaviour, dissent counts or the date of any term expiry; the listing is cited as a structural fact about who holds the seats.

6. Credit: FRED BAMLH0A0HYM2 at 3.30% and BAMLC0A0CM at 0.90%, both 11 November, against 3.42% and 0.93% on 6 November. Quoted for direction only; no all-in cost is computed in this piece.

7. Levels line: 2yr 4.86, 3yr 4.90, 5yr 4.95, 7yr 5.02, 10yr 5.10, 20yr 5.48, 30yr 5.43, 2s10s 24bp, 1y1y 5.07, all from the 12 November par close. data/latest.json carries a 12 November constant-maturity mark (UST10Y 5.103, UST30Y 5.433); those are a different series and are not mixed in.

CALL & INVALIDATION
Levels2Y 4.86 / 3Y 4.90 / 5Y 4.95 / 7Y 5.02 / 10Y 5.10 / 20Y 5.48 / 30Y 5.43 / 2s10s 24bp / 1y1y 5.07
CallThe political risk moved to 12 months and left the long end. The 1-year-ahead forward rose 38 basis points to 5.07% between the 6 and 12 November closes while the 5-year-ahead rose 17 to 5.25%, so the gap between them narrowed to -18 basis points from -45 - the 9th narrowest print of 217 sessions against a -59 median, with every other narrow print inside a 9-day September window. The long end did the opposite: the ACM 10yr term premium fell 21.1 basis points in the month to 6 November, to 0.8738% from 1.0844%, and the 30yr-minus-10yr spread at 33 basis points is the 19th narrowest of the file. The mechanism is a closed fiscal channel: a majority short of 60 Senate votes does not add duration supply, and the 4 November refunding held the 3yr, 10yr and 30yr sizes flat for a fourth quarter. What survives is the reaction function, which prices where an appointment can change it - at 12 months. The instrument is the 1y1y-minus-5y5y gap at -18, not the 10yr.
MechanismDivided government removes the fiscal channel and leaves the monetary one. Duration supply fixed means the premium for holding duration falls, which the term premium did by 21.1 basis points in a month. The policy rate is set regardless of who holds the House, so the surviving political variable is who runs the Fed and how fast a decision reverses; that variable prices at the horizon an appointment reaches, which is 12 to 24 months. The two legs are measured by different instruments and they moved in opposite directions in the same 6 sessions.
InvalidationThe 1y1y-minus-5y5y gap back at or below -45 basis points on a daily par close inside 10 sessions - 27 basis points of re-widening, through the whole November range, which would make the near-horizon move a positioning event rather than a policy-path repricing.
Second falsifierThe ACM 10yr term premium above 0.9857%, the 16 October print, on its next release: 11.2 basis points of re-widening from 0.8738%, which would say the fall was a wobble inside a 6-week build rather than a repricing of the fiscal channel.
Third falsifierThe 10yr breakeven more than 5 basis points from 2.24% on the next print. Above 2.29% puts part of the move back into inflation compensation; below 2.19% prices a growth problem instead of a policy one.
RELATED
The Vote Moved the 2yr 12bp and the 30yr 3bp. Only One of Them Is a Political Price. — the same front-versus-back split, measured on the session after the count
The Curve's Span Fell to a Year Low of 79 Basis Points. The 1yr Took 18 of Them in One Session. — the 1yr leg of this repricing, two sessions earlier
The Long End Has Already Voted. It Voted for Gridlock. — the term-premium level this piece marks down from
The State Cleared the Overnight Premium. The 2yr Rose 40bp Anyway. — the funding side of the same week, and why the state cannot set the 2yr
DISCLOSURE — NOT INVESTMENT ADVICE
NoticeMacroTape publishes macroeconomic analysis and market commentary for information and education only. Nothing published here is financial, investment, legal or tax advice, or a personal recommendation, and no author is a registered investment adviser. Calls, Levels and invalidation points are analytical views about how a market mechanism works, not instructions to trade. Readers act on their own responsibility and carry their own losses.