Macro ReviewHow to read
Macro · Policy & rates · Treasury curve

The 10-year Treasury yield has risen 47bp in three months

What is the Treasury curve pricing, and what moved the 10-year?

Treasury par yield curves · 16 Nov 2026FOMC decision 28 Oct 2026Term-premium models · 13 Nov 2026Updated 16 Nov 2026
10-year 5.12%16 Nov 2026 · 3m +47bp · since FOMC eve +13bp · 69th 1990
10-year real yield2.92%16 Nov 2026 · 3m +57bp · 100th pct since 2003
inflation2.20%16 Nov 2026 · 3m −10bp
+29bp10-year less 2-year · 16 Nov 2026 · 1m −6bpbear flattening
Slope, 3-month to 10-year (3m10y)+78bp16 Nov 2026 · 1m −5bppositive
5-year rate in five years ()5.37%16 Nov 2026 · 2015–19 mean 2.70%

The and slopes are the 10-year yield less the 2-year and less the 3-month bill; the rate is the five-year rate the curve locks in for five years from now.

In brief

What the curve says, in a few lines; what would change the read beneath
  • The 10-year is 5.12% on 16 Nov 2026, +98bp in a year.→ Exhibit 1
  • The 2-year to 10-year slope is +29bp, −6bp on the month: a .→ Exhibit 8
  • Of the 10-year's +47bp since 19 Aug 2026, +57bp was the real yield, not .→ Exhibit 5
  • The curve prices a one-year rate of 5.05% a year ahead, 115bp above the rate the Fed pays on reserves.→ Exhibit 10
  • Since the eve of the 28 Oct 2026 decision the 2-year is +13bp and the 10-year +13bp: a bear flattening.→ Exhibit 11
What would change this view
What would change this viewTriggerTodayDecided by
The month's bear flattening reverses > 32bp (halfway back to a month ago)+29bp (16 Nov 2026)not crossedeach Treasury close
The real-yield move reverses10-year real yield < 2.83% (halfway back)2.92% (16 Nov 2026)not crossedeach Treasury close
The curve stops pricing a rate above the forward < 3.90% ()5.05% (16 Nov 2026)not crossedeach Treasury close
The reasoning behind each point
  • Level
    The 10-year par yield is 5.12% on 16 Nov 2026, +98bp in a year.

    16 Nov 2026; the 69th percentile of all days since 1990. The 2-year is 4.83% and the 30-year 5.46%. Largest one-month move on the curve: the 1y (+16bp).

  • Shape
    The 2-year to 10-year slope is +29bp, −6bp on the month: a bear flattening.

    2y +9bp and 10y +3bp over 21 business days; 3m10y +78bp. 2s10s sits at the 28th percentile since 1990 (2015–19 mean 79bp); 5s30s +53bp, the 2s5s10s butterfly −9bp.

  • Real vs inflation
    Of the 10-year's +47bp since 19 Aug 2026, +57bp was the real yield, not inflation compensation.

    Real yields account for 121% of the move. 10-year real 2.92% (100th percentile since 2003); 10-year breakeven 2.20%, which also carries an inflation risk premium and the TIPS liquidity discount.

  • Forwards
    The curve prices a one-year rate of 5.05% a year ahead, 115bp above the rate the Fed pays on reserves.

    1y1y (the one-year Treasury rate one year ahead, read from today's curve) 5.05%; 2y1y 4.96%, +106bp over IORB (the rate the Fed pays on reserves, effective 17 Sep 2026). Treasury forwards include term premium and the Treasury–OIS basis, so when term premium is positive they sit above the expected policy path. The 5y5y nominal forward is 5.37% against a 2015–19 mean of 2.70%.

  • FOMC
    Since the eve of the 28 Oct 2026 decision the 2-year is +13bp and the 10-year +13bp: a bear flattening.

    From 27 Oct 2026 to 16 Nov 2026: 3m +12bp, 2y +13bp, 10y +13bp. IORB moved ±0bp with the decision. The previous decision was 16 Sep 2026; changes “since FOMC eve” in the tables count from the close before the latest decision.

  • Three months
    Of the 10-year's +21bp from 19 Aug 2026 to 6 Nov 2026, expected short rates carried +28bp on ACM.

    To 6 Nov 2026. ACM: expectations +28bp, term premium −13bp. Kim–Wright: expectations +11bp, term premium +4bp. Term premium is -82% of the fitted move on ACM and 25% on Kim–Wright.

  • Disagreement
    The two Fed models disagree on the 10-year term premium by 23bp on 6 Nov 2026.

    ACM − Kim–Wright; the 29th percentile of the gap since 1990. ACM is estimated on yields alone; Kim–Wright also uses survey forecasts of the 3-month bill rate, which anchors its expected path. Five to ten years ahead the expected short rate is 4.27% (ACM) and 3.98% (Kim–Wright).

The term premium · The read that the expected policy path drove the last three months fails if, over the next month, ACM's 10-year term premium rises above 0.79% and Kim–Wright's falls below 0.86% (half of the three-month move reversed) while the 10-year yield holds.

Next: Inside the 10-year — the premium, model by model

U.S. Department of the Treasury; Board of Governors (H.15, Policy Rates, Kim–Wright); Federal Reserve Bank of New York (ACM); Macro Review calculations · data to 16 Nov 2026rates.curve-read

The curve is +3bp at 10 years on the month: a

The par, zero, forward and real curves on selected dates, and the change split into real yields and

Exhibit 2 · The Treasury yield curve now and earlier, against the reserve rate

Nominal par yields, %; latest (16 Nov 2026), FOMC eve (27 Oct 2026), a month and a year earlier; dashed: IORB, 3.90% from 17 Sep 2026
About
About this exhibit
Nominal par yields at every published maturity (1 month to 30 years, log scale) on four dates: the latest close, the FOMC eve, a month earlier and a year earlier; dashed IORB line. Toggles switch to zero, 1-year forward or real par curves, and to changes.
The curve today, on the FOMC eve, a month and a year ago
Nominal par yields, every published tenor, %; the dashed line is IORB, 3.90% from 17 Sept 2026
Source: U.S. Department of the Treasury (par yields); Board of Governors (IORB); Macro Review calculations · data to 16 Nov 2026rates.curve-1

Exhibit 3 · Real and breakeven parts of changes in 5-year, 10-year and 30-year yields

Nominal yield change and real and breakeven parts, bp: a day, week, month, three months and year, and since the FOMC eve (27 Oct 2026).
About
Tenor1d1w1m3m1yFOMC eve
5yNominal−5+26+4+58+119+11
Real±0+26+16+61+132+31
Breakeven−5±0−12−3−13−20
10yNominal−3+24+3+47+98+13
Real+1+26+18+57+106+28
Breakeven−4−2−15−10−8−15
30yNominal−1+20+3+27+72+16
Real+2+22+14+36+79+25
Breakeven−3−2−11−9−7−9

Levels: 10-year 5.12% = real 2.92% + breakeven 2.20%. FOMC eve = close of 27 Oct 2026

About this exhibit
Table: change in the 5-, 10- and 30-year nominal par yields and their split into the real (TIPS) yield and the breakeven, bp, over one day, one week, one month, three months, a year and since the FOMC eve.
Nominal = real + inflation compensation
Change in the par yield and its split into the real yield and the breakeven, bp
Source: U.S. Department of the Treasury (nominal and real par yield curves); Macro Review calculations · data to 16 Nov 2026rates.curve-2

Exhibit 4 · The 10-year yield split into expected policy and term premium, monthly

New York Fed (ACM) model, %, month-end, Jun 1961 to Oct 2026; bars add to the line, the model's fitted 10-year zero-coupon yield
About this exhibit
Month-end stacked bars of the New York Fed (ACM) model's expected policy path and term premium, adding to its fitted 10-year yield (line); the whole record since Jun 1961 by default (periods: last ten years, All).
The 10-year: expected policy plus the term premium
% · month-end, the last ten years (All: since 1961) · the New York Fed's split of the fitted zero-coupon yield; the par yield differs by the coupon effect
Four models put the premium at 0.64–1.75% on 6 Nov 2026 (survey-anchored 1.75%, DKW 0.90%, Kim–Wright 0.87%, ACM 0.64%) — Inside the 10-year.
Source: Federal Reserve Bank of New York (ACM); Macro Review calculations · data to 13 Nov 2026rates.term_premium-11

Real yields carried 121% of the 10-year's three-month move

The nominal yield split two ways — into its real and breakeven parts, and into the and the on two Federal Reserve models

Exhibit 5 · The 10-year yield and its real and breakeven parts, daily

%, daily, 16 Nov 2016 to 16 Nov 2026; nominal par yield, real (TIPS) par yield and their gap, the breakeven; shaded: NBER recessions
About
About this exhibit
Daily 10-year nominal par yield, 10-year real (TIPS) par yield and 10-year breakeven, %, over the last ten years, recessions shaded, latest values marked.
10-year nominal, real and breakeven
%, 16 Nov 2016 to 16 Nov 2026
Source: U.S. Department of the Treasury (nominal and real par yield curves); NBER (recessions); Macro Review calculations · data to 16 Nov 2026rates.curve-7

Exhibit 6 · How the ACM and Kim–Wright models split the 10-year's change, by horizon

10-year change, bp, to 6 Nov 2026: one month, three months, a year and since the FOMC eve; bars add to each model's fitted 10-year.
About
About this exhibit
Stacked bars: change in the 10-year split into expected short rate and term premium on two Fed models (ACM, Kim–Wright), over one month, three months, a year and since the FOMC eve, to the latest day in both models; diamond = each model's fitted 10-year change.
Change in the 10-year yield, split by model, to 6 Nov 2026
bp; bars add to the diamond (each model's fitted zero-coupon yield). Changes to 6 Nov 2026, the latest day in both models.
Source: Federal Reserve Bank of New York (ACM); Board of Governors (Kim–Wright); Macro Review calculations · data to 13 Nov 2026rates.term_premium-1

Exhibit 7 · 10-year changes split into expectations and premium by model and horizon

10-year change, bp: one month, three months, a year and since the FOMC eve (27 Oct 2026); par and zero-coupon yields, then each model's split.
HorizonH.15 parGSW zeroACM exp.ACM TPKW exp.KW TP
1 monthfrom 16 Oct 2026−23−21−9−12−11−11
3 monthsfrom 19 Aug 2026+21+15+28−13+11+4
1 yearfrom 14 Nov 2025+72+66+65+2+32+37
Since FOMC evefrom 27 Oct 2026−13−10−12+3−5−6

“Since FOMC eve” starts at the close before the 28 Oct 2026 decision, the latest inside both files.

About this exhibit
Table: change in the 10-year over one month, three months, one year and since the FOMC eve: observed H.15 par and Board zero-coupon yields, then ACM and Kim–Wright expectations and term premium, bp (the same split as the chart beside it).
The same split as a table
bp; observed yields for reference
Source: Federal Reserve Bank of New York (ACM); Board of Governors (Kim–Wright, staff curve, H.15); Macro Review calculations · data to 13 Nov 2026rates.term_premium-2

The depends on the path of short rates each model expects: four models put it at 0.64–1.75% on 6 Nov 2026. The split into four parts, the band and what drives it: Inside the 10-year.

Is the curve steep or flat, and the premium high or low, against history?

The curve's slopes since 1990 and the since 1961

Exhibit 8 · The Treasury curve's slope by span, daily

Longer less shorter yield, bp, daily, 16 Nov 2016 to 16 Nov 2026: 2-year to 10-year, 3-month to 10-year, 5-year to 30-year; shaded: NBER recessions
About
About this exhibit
Daily 2-year to 10-year (2s10s), 3-month to 10-year (3m10y) and 5-year to 30-year (5s30s) slopes, bp, over the last ten years, zero line, recessions shaded.
The curve's slope at three spans
bp, 16 Nov 2016 to 16 Nov 2026
Source: U.S. Department of the Treasury (par yields); NBER (recessions); Macro Review calculations · data to 16 Nov 2026rates.curve-5

Exhibit 9 · The 10-year term premium on the ACM and Kim–Wright models since 1961

%, month-end, Jun 1961 to Oct 2026; ACM (New York Fed) from 1961, Kim–Wright (Board) from 1990; shaded: NBER recessions
About this exhibit
Month-end 10-year term premium on the New York Fed's ACM model (from Jun 1961) and the Board's Kim–Wright model (from 1990), %, recessions shaded.
10-year term premium since 1961 (month-end)
%, Jun 1961 to Oct 2026
Source: Federal Reserve Bank of New York (ACM); Board of Governors (Kim–Wright); NBER (recessions); Macro Review calculations · data to 13 Nov 2026rates.term_premium-5

What policy path does the curve price?

The one-year rates the curve locks in against the rate the Fed pays on reserves

Exhibit 10 · Forward rates the Treasury curve locks in, against the reserve rate

%, daily, last five years to 16 Nov 2026; 1-year a year ahead (1y1y) and two years ahead (2y1y), 5-year five years ahead (5y5y); dashed: IORB
About
About this exhibit
Daily forward rates bootstrapped from the par curve (one-year rate a year ahead 1y1y, one-year rate in two years 2y1y, five-year rate in five years 5y5y) with IORB (dashed), %, over the last five years by default.
The curve prices one-year rates above the rate on reserves
%, 16 Nov 2021 to 16 Nov 2026
Source: U.S. Department of the Treasury (par yields); Board of Governors (IORB); Macro Review calculations · data to 16 Nov 2026rates.curve-8

Every tenor, spread and forward, the shapes against history, and the bootstrap's accuracy

The full ladder of yields, shapes and forwards with a chart of the selected row, each shape against its own distribution, the two models' daily detail, and the leave-one-out test of the bootstrap

Exhibit 11 · Levels and changes of every Treasury yield, slope and forward

%, latest level on 16 Nov 2026; change (bp) over a day, week, month, three months and year, and since the FOMC eve; percentile since 1990.
SeriesLevel1d1w1m3m1ySince FOMC evePct since 1990Level, recentAs of
Nominal par yields
1-month par yield4.10%±0bp+6bp−2bp+33bp+6bp±0bp8216 Nov 2026
3-month par yield4.34%+2bp+10bp+8bp+48bp+39bp+12bp6616 Nov 2026
6-month par yield4.42%±0bp+16bp+8bp+48bp+62bp+8bp6616 Nov 2026
1-year par yield4.62%−3bp+24bp+16bp+62bp+92bp+12bp6816 Nov 2026
2-year par yield4.83%−5bp+27bp+9bp+64bp+121bp+13bp7316 Nov 2026
3-year par yield4.87%−5bp+29bp+5bp+62bp+126bp+7bp7316 Nov 2026
5-year par yield4.93%−5bp+26bp+4bp+58bp+119bp+11bp7016 Nov 2026
7-year par yield5.01%−4bp+25bp+3bp+53bp+109bp+10bp6916 Nov 2026
10-year par yield5.12%−3bp+24bp+3bp+47bp+98bp+13bp6916 Nov 2026
20-year par yield5.51%−1bp+22bp+4bp+34bp+78bp+17bp7516 Nov 2026
30-year par yield5.46%−1bp+20bp+3bp+27bp+72bp+16bp6616 Nov 2026
Curve shape
2s10s (10-year − 2-year)29bp+2bp−3bp−6bp−17bp−23bp±0bp2816 Nov 2026
About this exhibit
Table of the curve's series (par yields 1 month to 30 years, curve slopes, real yields, breakevens, forwards, IORB): latest level, change over one day, one week, one month, three months, one year and since the FOMC eve, percentile since 1990, sparkline; picking a row charts it in the chart at the top.
Treasury yields, shapes and forwards
Latest level; changes over one day, one week, one month, three months, a year (from the close on or before the same day a year earlier) and since the close before the latest FOMC decision. Change cells are shaded against each series' own 2015–19 distribution. Select a row to chart it.
Source: U.S. Department of the Treasury (par yields); Board of Governors (H.15, IORB); Macro Review calculations · data to 16 Nov 2026rates.curve-3

Exhibit 1 · The 10-year Treasury par yield over time

Par yield, %, daily, last five years to 16 Nov 2026; dot: the latest close; pick a row in the yields table to chart another series
About this exhibit
The daily 10-year Treasury par yield, last five years by default (1y/5y/10y/Max buttons), latest close marked; picking a row in the yields table redraws and retitles it for that series.
10-year par yield: 5.12% on 16 Nov 2026
%, 16 Nov 2021 to 16 Nov 2026
Source: U.S. Department of the Treasury; Board of Governors; Macro Review calculations · data to 16 Nov 2026rates.curve-4

Exhibit 12 · Curve measures against their history since 1990 or 2003

Value (% or bp) on 16 Nov 2026 and its percentile against every published day since the start shown.
About
MeasureSinceTodayPctMinp5p25Medianp75p95Max2015–19 meanDays < 0
2s10s2 Jan 1990+29bp28−108bp2023−35bp+25bp+83bp+176bp+251bp+291bp2011+79bp11%
5s30s2 Jan 1990+53bp36−61bp2000−1bp+38bp+80bp+142bp+230bp+302bp2010+91bp5%
3m10y2 Jan 1990+78bp33−189bp2023−74bp+51bp+145bp+250bp+345bp+387bp1992+119bp12%
2s5s10s butterfly2 Jan 1990−9bp33−79bp2010−47bp−16bp+2bp+22bp+51bp+82bp1992+4bp47%
10-year par yield2 Jan 19905.12%690.52%20201.56%2.64%4.20%5.67%7.80%9.09%19902.27%
10-year real yield2 Jan 20032.92%100-1.19%2021-0.84%0.31%1.08%1.90%2.41%3.15%20080.48%
10-year breakeven2 Jan 20032.20%470.04%20081.45%1.87%2.23%2.36%2.57%3.02%20221.79%
5y5y nominal forward2 Jan 19905.37%650.62%20201.99%3.28%4.84%5.91%8.47%9.65%19902.70%
5y5y forward breakeven2 Jan 20032.17%270.48%20081.71%2.15%2.34%2.57%2.94%3.54%20082.00%
About this exhibit
Table: the curve measures (2s10s, 5s30s, 3m10y, butterfly, 10-year nominal, real and breakeven, 5y5y nominal forward and forward breakeven): value on the latest close and its percentile against every day since 1990 or 2003, with min, p5, p25, median, p75, p95, max, 2015–19 mean and share of days below zero.
Distribution since the series began
Today against every published day since the start shown; p5–p95 spans the middle 90% of days
Source: U.S. Department of the Treasury (par yields: nominal since 1990, real since 2003); Macro Review calculations · data to 16 Nov 2026rates.curve-6

Exhibit 13 · How closely the site's zero curve reproduces published yields, by maturity

Leave-one-out: drop one published tenor, fit the curve again, compare its implied par yield with the published one (bp, to 16 Nov 2026).
Dropped tenorMedian95th pctMax
2y4.89.112.2
3y1.54.57.2
5y1.74.04.7
7y2.54.15.5
10y3.85.36.1
20y16.219.219.8
Against the Board staff curve5y10y30y
Zero − par, this curve (median, bp)1.05.415.4
Zero − par, GSW (median, bp)1.16.319.5
Gap in zero − par (max, bp)0.51.416.3
Zero-rate level gap (median, bp)on-the-run vs off-the-run fit2.06.08.2
Correlation of daily changes0.980.970.81

The Board staff curve (Gürkaynak, Sack and Wright) fits a Svensson curve to off-the-run notes and bonds, so its level differs from Treasury's on-the-run par curve (10-year zero: median gap 6.0bp over the check window). The coupon effect — zero minus par at the same maturity — is a property of the curve's shape, not of which bonds it is fitted to, and is what the site's bootstrap is checked on. The staff file runs to 6 Nov 2026.

About this exhibit
Table: leave-one-out error (median, 95th percentile, max, bp, recent days) when each published tenor is dropped and the curve fitted again, plus coupon-effect and level gaps and daily-change correlations against the Board staff curve at 5, 10 and 30 years.
How accurate is the bootstrap?
Leave-one-out: drop one published tenor, refit the curve, compare its implied par yield with the published one (last 250 days, bp)
Source: U.S. Department of the Treasury (par yields); Board of Governors (staff zero-coupon curve); Macro Review calculations · data to 16 Nov 2026rates.curve-10

Exhibit 14 · The 10-year term premium on the ACM and Kim–Wright models, daily

%, daily, 14 Nov 2016 to 13 Nov 2026; ACM (New York Fed) to 13 Nov 2026, Kim–Wright (Board) to 6 Nov 2026
About
About this exhibit
Daily 10-year term premium on ACM and Kim–Wright, each to its own file's latest day, %, over the last ten years.
10-year term premium
%, 14 Nov 2016 to 13 Nov 2026
Source: Federal Reserve Bank of New York (ACM); Board of Governors (Kim–Wright); Macro Review calculations · data to 13 Nov 2026rates.term_premium-3

Exhibit 15 · The ACM and Kim–Wright expected short rate: ten-year and five-to-ten-year averages

%, daily, 14 Nov 2016 to 13 Nov 2026; solid: average over the next ten years; dashed: average five to ten years ahead
About
About this exhibit
Daily expected short rate on ACM and Kim–Wright: the average over the next ten years (solid) and the average five to ten years ahead (dashed), %, over the last ten years.
Expected short rate, 10-year average and 5–10 years ahead
%, 14 Nov 2016 to 13 Nov 2026
Source: Federal Reserve Bank of New York (ACM); Board of Governors (Kim–Wright); Macro Review calculations · data to 13 Nov 2026rates.term_premium-4

Exhibit 16 · Gap between the ACM and Kim–Wright 10-year term premia, daily

ACM less Kim–Wright, bp, daily, 14 Nov 2016 to 6 Nov 2026; below zero: ACM's premium is the lower
About this exhibit
Daily gap between the ACM and Kim–Wright 10-year term premia (ACM less Kim–Wright), bp, over the last ten years, zero line; legend reads 'Gap'.
The two models' premia apart, ACM less Kim–Wright
bp, 14 Nov 2016 to 13 Nov 2026
Source: Federal Reserve Bank of New York (ACM); Board of Governors (Kim–Wright); Macro Review calculations · data to 13 Nov 2026rates.term_premium-8

Exhibit 17 · How far each Fed model's fitted 10-year misses the observed yield

Fitted less observed 10-year zero-coupon yield (Board staff curve), bp, daily, last five years to 6 Nov 2026; ACM and Kim–Wright
About
About this exhibit
Daily gap between each model's fitted 10-year yield and the observed Board staff 10-year zero-coupon yield, ACM and Kim–Wright, bp, over the last five years by default.
Fitted minus observed 10-year zero-coupon yield
bp, 15 Nov 2021 to 13 Nov 2026
Source: Federal Reserve Bank of New York (ACM); Board of Governors (Kim–Wright, staff curve); Macro Review calculations · data to 13 Nov 2026rates.term_premium-9

Exhibit 18 · The ACM and Kim–Wright 10-year split: levels and recent changes

%, latest level; change (bp) over a day, week, month, three months and year, and since the FOMC eve; percentile since 1990.
SeriesLevel1d1w1m3m1ySince FOMC evePct since 1990Level, recentAs of
Observed
10-year Treasury par yield (H.15)5.15%+5bp+29bp+5bp+44bp+115bp+16bp7013 Nov 2026
10-year zero-coupon yield (Board staff curve)continuously compounded; the yield both models fit4.88%−1bp−6bp−36bp+9bp+68bp−10bp646 Nov 2026
ACM model (New York Fed)
ACM fitted 10-year yield5.11%+5bp+25bp+4bp+33bp+106bp+16bp6713 Nov 2026
Expected average short rate (risk-neutral yield)4.39%±0bp+17bp+6bp+47bp+89bp+4bp7213 Nov 2026
Term premium0.73%+5bp+8bp−3bp−14bp+17bp+12bp4313 Nov 2026
Kim–Wright model (Board)
Kim–Wright fitted 10-year yield4.89%+1bp−4bp−36bp+12bp+70bp−11bp646 Nov 2026
Expected average short ratefitted yield − term premium4.01%+1bp−1bp−16bp+9bp+33bp−5bp636 Nov 2026
Term premium0.87%±0bp−4bp−20bp+3bp+37bp−6bp566 Nov 2026
Term premium by maturity
ACM 2-year0.33%+2bp+19bp±0bp+13bp+33bp+6bp5513 Nov 2026
ACM 5-year0.43%+4bp+10bp−2bp−3bp+30bp+8bp4713 Nov 2026
Kim–Wright 2-year0.40%±0bp−4bp−12bp+4bp+31bp−6bp666 Nov 2026
Kim–Wright 5-year0.57%±0bp−5bp−17bp+4bp+40bp−7bp646 Nov 2026
About this exhibit
Table of the models' series: observed 10-year par and zero-coupon yields; ACM and Kim–Wright fitted 10-year yield, expected average short rate and term premium; term premia at 2 and 5 years; latest level, changes over six horizons, percentile since 1990; picking a row charts it in the chart below.
Term structure model outputs
Latest level; changes in bp over one day, one week, one month, three months, a year and since the eve of the latest FOMC decision in both files. Each row runs to its own file's latest day. Select a row to chart it.
Source: Federal Reserve Bank of New York (ACM); Board of Governors (Kim–Wright, staff curve, H.15); Macro Review calculations · data to 13 Nov 2026rates.term_premium-6

Exhibit 19 · The New York Fed model's 10-year term premium over time

%, daily, last five years to 13 Nov 2026; shows the row picked in the model table
About this exhibit
The row picked in the model table, daily, over the last five years by default; by default the ACM 10-year term premium (retitled, e.g. 'ACM 5-year', when another row is picked).
ACM model (New York Fed) — Term premium
%, 15 Nov 2021 to 13 Nov 2026
Source: Federal Reserve Bank of New York (ACM); Board of Governors; Macro Review calculations · data to 13 Nov 2026rates.term_premium-7

Appendix

Where the curve comes from, how the bootstrap derives zeros and forwards, and which models split the 10-year70 checks, all pass
About this page

The Treasury curve is the price of money over every horizon: the front end follows the expected policy path, the long end adds compensation for inflation and for holding duration. This page reads Treasury's own par curve, splits yields into real and inflation compensation, splits the 10-year's move into expected policy and the premium on two Federal Reserve models, and bootstraps the forwards the curve locks in.

Exhibit 20 · Treasury curve: definitions

  • Source
    Treasury's par curve is H.15.

    Treasury's par yield curve rates are the H.15 constant-maturity yields: the Board republishes Treasury's numbers, and the checks confirm it for every common day (nominal since 1990, TIPS since 2003). Treasury reads par yields off a curve fitted to indicative bid-side closing prices of the most recently auctioned securities — a monotone convex spline since 6 Dec 2021, a quasi-cubic Hermite spline before. Bill yields are bond-equivalent.

  • Calendar
    Business days, gaps kept.

    Business days (Mon–Fri); a one-month change always spans 21 business days. Days Treasury does not publish (market holidays: 396 since 1990) stay empty in the level series; a change that would start on one is measured from the previous close (carried forward at most three days, never across the 2002–06 30-year gap). This page holds the last ten years of daily levels; percentiles and distributions use every published day since the series began.

  • Breaks
    Tenors come and go.

    The 30-year was not published from 19 Feb 2002 to 8 Feb 2006 (5s30s is blank there). The 20-year starts 1 Oct 1993; the 1-month 31 Jul 2001; the 2-month 16 Oct 2018; the 4-month 19 Oct 2022; the 6-week 18 Feb 2025. Real par yields start 2 Jan 2003 (20-year 27 Jul 2004, 30-year 22 Feb 2010).

  • Bootstrap
    Par → discount factors → zero and forward rates.

    Par yields are interpolated to a semi-annual grid with a monotone cubic Hermite spline (PCHIP) in maturity. Maturities up to six months are single-payment bonds (D = 1 / (1 + y·t)); longer par bonds pay y/2 each half-year and price at par, which gives each discount factor in turn: D(tₙ) = (1 − y/2 · ΣD(tᵢ<tₙ)) / (1 + y/2). Zero and forward rates are quoted semi-annually compounded. Nothing is extrapolated beyond the longest tenor published that day.

  • Real curve
    Five years and out.

    The real curve starts at five years. It is extended flat to the left from the 5-year real par yield before bootstrapping; a flat par curve is also a flat zero curve, so the 5-year real zero equals its par yield and the assumption shapes only the interpolation. The 5y5y forward breakeven is the nominal 5y5y forward minus the real 5y5y forward; two-decimal par yields put roughly ±3bp of rounding noise into it day to day.

  • Forwards
    What the curve locks in.

    f(t₁, t₂) = 2 · ((D(t₁) / D(t₂))^(1/(2(t₂ − t₁))) − 1). Forward rates are what the Treasury curve locks in for a future period; they include term premium, so a forward above the policy rate is not by itself a forecast of hikes. The 10y10y relies on interpolation between the 10- and 20-year par yields — the leave-one-out error at 20 years is shown above.

  • Breakevens
    Compensation, not expectation.

    Breakeven = nominal par yield − real par yield at the same maturity. It is inflation compensation, not an inflation forecast: it includes an inflation risk premium and the liquidity discount of TIPS, and real par yields are on CPI-indexed principal with a three-month indexation lag.

  • FOMC
    Changes since the decision eve.

    FOMC decision dates are the policy-statement dates on the Board's FOMC calendar. “Since FOMC eve” counts from the close of the business day before the latest decision (27 Oct 2026, before 28 Oct 2026); Treasury's par yields are end-of-day readings, so a decision-day move includes the statement and press conference.

  • Norms
    2015–19 shading, full-history percentiles.

    Shading compares each change with its own distribution over 2015–2019; percentiles run since 1990 for nominal yields, shapes and nominal forwards and since 2003 for real yields and breakevens (30-year real since Feb 2010).

  • IORB
    A reference line.

    IORB (IOER before 29 Jul 2021) from the Board's policy-rate release is shown only as a reference point for the forwards. The Board dates a new setting from its effective day, the day after the decision.

Source: U.S. Department of the Treasury (par yield curve csv); Board of Governors (H.15, Policy Rates, staff yield curves, FOMC calendar); Macro Review calculations · data to 16 Nov 2026rates.curve-9

Exhibit 21 · Term premia: definitions

  • Models
    Two affine term structure models.

    ACM (Adrian, Crump and Moench 2013) is a five-factor affine model estimated by regressions on zero-coupon yields alone; the New York Fed publishes daily and month-end fitted yields, risk-neutral yields (the expected average short rate) and term premia for 1–10 years. Kim–Wright (2005) is a three-factor affine model that also uses survey forecasts of the 3-month bill rate; the Board publishes fitted yields, term premia and forward rates, and defines the expectations component as fitted yield minus term premium.

  • Definition
    Yield = expectations + premium.

    Term premium = zero-coupon yield − expected average short rate over the life of the bond. Kim–Wright's term premium includes the (negative, small) convexity premium; ACM's risk-neutral yield embeds convexity the same way, so the two are comparable in changes and close in levels.

  • Inputs
    Fitted to the zero-coupon curve.

    Both models are fitted to the Board staff zero-coupon curve (Gürkaynak, Sack and Wright), which is fitted to off-the-run notes and bonds. That is why their fitted 10-year yield sits a few basis points from the on-the-run H.15 par yield: the coupon effect (zero vs par) plus the on-the-run premium.

  • Forward
    Five to ten years ahead.

    The 5-to-10-year expected short rate and forward term premium use continuous compounding: f(5,10) = 2 × y(10) − y(5), applied to each component.

  • Calendar
    Business days; files lag.

    Business days. Days a file does not publish stay empty; a change starting on one is measured from the previous close. ACM runs to 13 Nov 2026 and Kim–Wright to 6 Nov 2026; the change decomposition uses the common end date 6 Nov 2026. This page holds ten years daily and the monthly history since 1961 (ACM) and 1990 (Kim–Wright).

  • FOMC
    The latest decision both files cover.

    “Since FOMC eve” counts from the close of 27 Oct 2026, the business day before the 28 Oct 2026 decision — the latest decision inside both files.

  • Caveats
    Estimates, not prices.

    Term premia are model estimates, not market prices: they depend on the model's factors, sample and (for Kim–Wright) the survey data. Both are revised when re-estimated. Read changes over weeks and months rather than single days; the disagreement between the models is itself a measure of uncertainty.

  • Licence
    Attribution.

    ACM data © Federal Reserve Bank of New York, used with attribution; the New York Fed is not affiliated with this site. Kim–Wright and GSW are Board staff research data, not official statistical releases.

  • Norms
    2015–19 shading.

    Shading compares each change with its own 2015–19 distribution; percentiles run since 1990 for both models.

Source: Federal Reserve Bank of New York (ACM); Board of Governors (Kim–Wright model page and Q&A, staff curve, H.15); Macro Review calculations · data to 13 Nov 2026rates.term_premium-10

Before you read

Macro Review is a personal research site, still being built.

Nothing on it is investment advice, a recommendation or a solicitation.

Check every figure for yourself. Each names its publisher — confirm it there before you rely on it.

Figures can be revised, delayed or wrong.

Cones, fans, scenarios, fair values and “what followed” tables are conditional on the model each page names: they describe, they do not predict.

No warranty is given; use is at your own risk.