Prolepsia forecastYes / no

Will the gap between 10-year and 2-year US Treasury yields be above 75 basis points on 31 December 2027?

As of 15 November 2026, Prolepsia puts the chance of yes at 30%.

30%chance of yes

Yes 30%No 70%
Prolepsia forecast made

How Prolepsia reasons

The report Prolepsia wrote with this forecast.

How forecasts are made · Track record

TL;DR

The saved forecast is 30.3% Yes and 69.7% No. The more likely outcome is a spread at or below 75 basis points at the specified year-end observation, because the curve starts relatively flat and current inflation and policy evidence do not support assuming rapid easing.

Key evidence

  • The starting gap is substantial. On November 13, 2026, Treasury reported a 10-year yield of 5.15% and a 2-year yield of 4.88%, giving a 27-basis-point spread. It must widen by more than 48 basis points to qualify; exactly 75 resolves No. (Treasury rates)
  • Historical flat-curve episodes favor No. For starting spreads between zero and 50 basis points, the share above 75 basis points fourteen months later was 25% over the full history, 20% after 1990, and 23% after 2008. These are historical calculations, not independent trials: nearby observations often belong to the same economic cycle. (FRED historical series)
  • Policy remains restrictive. The October 28 FOMC maintained a 3.75%–4.00% target range, with two dissenters favoring an increase. October headline CPI rose 0.4% month over month and core CPI rose 0.3%. That combination argues against treating an imminent easing cycle as the default. (FOMC statement; CPI release)
  • Supply does not establish an automatic steepening catalyst. Treasury’s November refunding guidance kept near-term coupon auction sizes unchanged. (Treasury refunding statement)

Uncertainties

A growth downturn could prompt easing and pull the two-year yield down faster than the ten-year yield. Higher compensation for holding longer-term debt could also widen the spread without rate cuts. Conversely, persistent inflation or renewed tightening could keep the curve flat or inverted.

Historical experience spans different monetary regimes and relatively few independent episodes. The current difference between long- and short-maturity term premia remains unverified. Finally, this is a single-date forecast: crossing the threshold during 2027 is insufficient, and late-year market moves can determine the result.

Read the full report

TL;DR

The saved forecast is 30.3% Yes and 69.7% No. The forecast favors a 10-year minus 2-year Treasury spread at or below 75 basis points at the specified year-end observation. The decisive distinction is the endpoint: an earlier crossing does not satisfy the question.

Context

As of the November 15, 2026 forecast date, the latest Treasury observation was November 13: the ten-year yield was 5.15% and the two-year yield was 4.88%, leaving a spread of 27 basis points. The spread therefore needs more than 48 basis points of widening to exceed the threshold. At the current two-decimal yield reporting precision, the first qualifying published spread is 76 basis points. (Treasury daily rates)

Resolution uses the official Treasury par yields for December 31, 2027, or the most recent preceding publication day if necessary. Exactly 75 basis points resolves No. These are same-date par-curve observations, not intraday quotes or yields on particular outstanding securities. The designated fallback is the FRED T10Y2Y series.

Evidence

The historical backbone favors No. In the history beginning in June 1976, starting spreads between zero and 50 basis points finished above 75 basis points fourteen months later in 25% of observations. The corresponding frequencies were 20% after 1990 and 23% after 2008. Fourteen months is a close approximation to the remaining horizon. These calculations describe what followed similarly flat curves, rather than how often a steep curve existed in general. (FRED historical spread data)

That distinction matters. Across complete monthly observations through October 2026, the spread exceeded the threshold roughly half the time over the full history. But today’s starting point is well below the threshold, and yield-curve regimes persist. An unconditional historical frequency answers a different question from the one facing the client. The historical record also contains long flat-curve stretches and concentrated steepening episodes, so adjacent monthly observations are not independent evidence. (FRED historical spread data)

The recent history reinforces the low starting point. The full sequence available from December 2025 through October 2026 is below, followed by the latest partial-November observation. All spreads are in percentage points; the November entry is not a completed month-end. (FRED spread history)

Observation Spread, percentage points
December 2025 month-end 0.71
January 2026 month-end 0.74
February 2026 month-end 0.59
March 2026 month-end 0.51
April 2026 month-end 0.52
May 2026 month-end 0.47
June 2026 month-end 0.30
July 2026 month-end 0.47
August 2026 month-end 0.41
September 2026 month-end 0.41
October 2026 month-end 0.38
November 13, 2026 0.27

Every completed 2026 month-end in that sequence was below the threshold. The latest weekly move also shows why higher bond yields do not automatically imply steepening. On November 6, the ten-year and two-year yields were 4.86% and 4.51%, respectively. By November 13, both had risen, but the two-year rose more, reducing the spread from 35 to 27 basis points. (Treasury daily rates)

The policy backdrop supports caution about rapid easing. On October 28, the FOMC retained a 3.75%–4.00% target range, described strong activity and elevated inflation, and recorded two dissents favoring a quarter-point increase. In his November 10 speech, Governor Waller favored further tightening and said a half-point December increase should be considered if monthly core inflation remained around 0.3% or higher. His view is not a committee commitment, but it is evidence against assuming that cuts are imminent. (FOMC statement; Waller speech)

October inflation provided no clear release from that constraint. Headline CPI rose 0.4% month over month and core CPI rose 0.3%, both seasonally adjusted. The usual annual comparisons were unavailable because the relevant October 2025 observations were missing. No unofficial annual estimate is needed to draw the narrower conclusion: the latest monthly readings did not establish an imminent easing cycle. (BLS CPI release)

The September policy projections also leaned against large cuts as the central path. Participants’ median year-end federal-funds projection was 4.1% for both 2026 and 2027, while median PCE inflation declined from 3.7% to 2.3%. These are conditional policy assessments, not promises or Treasury-spread forecasts. They nevertheless show that projected disinflation did not require a sharply lower year-end policy rate. (September projections)

Labor evidence is mixed rather than recessionary across the board. October payrolls increased by 165,000, unemployment was 4.1%, and temporary-help employment rose by 18,600. Against that, involuntary part-time employment increased by 339,000 to 4.84 million, including a rise of 356,000 attributed to slack work or business conditions. The latter is a warning about hidden weakness, but one volatile household-survey month does not establish a downturn. (October employment release)

Bank lending evidence tells a similar story. In the October lending survey, more banks reported weaker than stronger business-loan demand. Yet inquiries about new or larger credit lines rose at nine banks, fell at seven, and were unchanged at 44. The inference is softness without a demonstrated collapse in the financing pipeline. (Senior Loan Officer Opinion Survey)

Supply and central-bank holdings offer less support for automatic steepening than a simple deficit narrative suggests. Treasury’s November refunding guidance maintained near-term coupon sizes and placed marginal seasonal financing changes in bills. Advice to consider coupon increases during April–September 2027 was not an adopted schedule, and its proposed concentration in the two- through ten-year sector would affect both sides of this spread. (Treasury refunding statement)

Nor was ongoing Treasury balance-sheet runoff the operative policy. The October implementation directive rolled over all Treasury principal and reinvested agency principal into Treasury bills. That removes a simple argument that continued Treasury runoff must push longer yields higher. Bill reinvestment is also different from buying long-duration bonds. (Federal Reserve implementation note)

There is still a meaningful long-end demand risk. Treasuries held in Federal Reserve custody for foreign official accounts stood at $2.531 trillion in the week ended November 11, down $217.726 billion year over year and $19.076 billion week over week. That supports watching foreign demand, but custody holdings are not total foreign ownership. Their decline cannot be equated with the same amount of net selling. (Federal Reserve balance-sheet release)

Short-term funding and auction evidence do not establish a separate structural shock. The reviewed funding window showed limited rate pressure, while comparable primary auction history was insufficient to establish a persistent deterioration in demand. A high-minus-median auction yield difference is not the same as an auction tail against the when-issued market. Those distinctions prevent isolated figures from carrying more weight than they deserve. (SOFR data; November ten-year auction result)

What's non-obvious

The spread is a relative price, not a forecast of whether rates rise or fall. A selloff can flatten the curve, as the latest weekly observations show. Rate cuts can also fail to deliver enough steepening if long yields fall alongside short yields. The two-year yield already stands well above the current policy range, so a retreat in expected restriction could widen the curve without an outright recession. That is an inference from the starting configuration, not a uniquely identified policy path.

The other overlooked point is that substantial normalization can still resolve No. TD Economics projected fourth-quarter 2027 yields of 4.35% for ten years and 3.65% for two years, implying a 70-basis-point spread. That institutional projection is not the saved forecast, and a quarter-end outlook is not a guaranteed daily observation. It illustrates the narrow distinction: a much steeper curve than today can still finish below the required threshold. (TD Economics forecast tables)

Uncertainties

The largest risk to No is a material downturn that drives the two-year yield down faster than the ten-year yield. Orderly disinflation could do the same on a smaller scale. A rise in compensation for holding longer maturities provides another route to Yes, even without policy easing. Persistent inflation and renewed tightening point the other way.

The current ten-year-minus-two-year term-premium differential remains unverified. Dated estimates for both maturities would help distinguish expected-policy effects from changes in duration compensation. Broader, comparable primary auction history would also clarify whether demand weakness is persistent rather than isolated. (New York Fed term-premium information)

Historical evidence is limited by regime changes and clustering. The record spans very different inflation and policy environments, while many apparently separate observations repeat the same episodes. It is also current-vintage history, not a reconstruction of the information available at each historical date. These limits weaken claims of precision; they do not overturn the direction of the forecast.

Finally, the outcome depends on one official observation. Late-year repricing, publication-day fallback rules, and revisions before resolution matter most near the boundary. The saved percentages express the forecast, not certainty to a tenth of a percentage point.

Forecast history

How Prolepsia’s forecast has movedOne forecast so far: 30% on 15 Nov 2026.
DateForecast
30% chance of yes

Sources Prolepsia read

  1. Daily Treasury Rates | U.S. Department of the Treasuryhome.treasury.gov
  2. 10-Year Treasury Constant Maturity Minus 2-Year Treasury Constant Maturity (T10Y2Y) | FRED | St. Louis Fedfred.stlouisfed.org
  3. Federal Reserve Board - Federal Reserve issues FOMC statementwww.federalreserve.gov
  4. Consumer Price Index Summary - 2026 M10 Resultswww.bls.gov
  5. Secured Overnight Financing Rate Data - FEDERAL RESERVE BANK of NEW YORKwww.newyorkfed.org
  6. November ten-year auction resultwww.treasurydirect.gov
  7. TD Economics - U.S. Latest Forecast Tableseconomics.td.com

The question

This question asks whether the spread between the 10-year and 2-year U.S. Treasury yields will exceed 75 basis points (0.75 percentage points) on December 31, 2027. The spread is calculated by subtracting the 2-year Treasury yield from the 10-year Treasury yield, using the U.S. Department of the Treasury's daily par yield curve rates. A positive spread indicates that longer-term Treasury securities yield more than shorter-term securities, while a negative spread indicates an inverted yield curve. As of November 13, 2026, the latest available trading day before this question was specified, the 10-year Treasury yield was approximately 5.15% and the 2-year yield was approximately 4.88%, producing a spread of 27 basis points. The spread would therefore need to widen by more than 48 basis points from that level to exceed the specified threshold. The spread is influenced by expectations about Federal Reserve monetary policy, inflation, economic growth, and the term premium demanded by investors for holding longer-term Treasury securities. The question concerns the yield spread on December 31, 2027, rather than its average during 2027 or whether it exceeds 75 basis points at any earlier point. Background sources include the U.S. Treasury's Daily Treasury Par Yield Curve Rates (https://home.treasury.gov/resource-center/data-chart-center/interest-rates) and the Federal Reserve Bank of St. Louis's FRED series T10Y2Y (https://fred.stlouisfed.org/series/T10Y2Y).

How it resolves

The question resolves YES if the difference between the 10-year and 2-year U.S. Treasury yields on December 31, 2027, is strictly greater than 75 basis points (0.75 percentage points). It resolves NO if the difference is less than or equal to 75 basis points. The primary resolution source will be the U.S. Department of the Treasury's official Daily Treasury Par Yield Curve Rates, specifically the published 10-year and 2-year par yields for December 31, 2027. The spread will be calculated as (10-year yield minus 2-year yield) × 100, with both yields expressed as percentages. For example, a 10-year yield of 4.50% and a 2-year yield of 3.70% would produce a spread of 80 basis points and resolve YES. A 10-year yield of 4.50% and a 2-year yield of 3.75% would produce a spread of exactly 75 basis points and resolve NO. If the Treasury does not publish yields for December 31, 2027, because of a market holiday, closure, or another reason, the observation from the most recent preceding business day for which both yields are officially available will be used. If the Treasury's data are permanently unavailable, the fallback source will be the Federal Reserve Bank of St. Louis's FRED series T10Y2Y, which reports the 10-year minus 2-year Treasury constant-maturity spread in percentage points. The latest available observation on or before December 31, 2027, will be used. If neither source provides a usable observation for the relevant period, the question will be annulled. Resolution will use the official published values, without substituting intraday market quotes, yields on individual Treasury securities, or alternative estimates. The outcome will be determined once the relevant official data are available, allowing for routine publication delays.

Fine print

The threshold is strictly greater than 75 basis points; a spread of exactly 75 basis points does not qualify. One basis point equals 0.01 percentage points, so 75 basis points equals 0.75 percentage points. Both yields must refer to the same observation date and the same methodology. The U.S. Treasury's published par yield curve rates are used rather than the yields of specific outstanding Treasury notes, which may differ depending on their remaining maturities and market prices. The spread is measured at a single observation date, not averaged over a period. The final published Treasury values for the relevant observation date will take precedence over preliminary figures or third-party estimates. Subsequent revisions published before resolution will be incorporated; revisions made after the question has been resolved will not ordinarily reopen the outcome. If December 31, 2027, is not an official Treasury yield publication day, the fallback to the preceding available business day applies automatically. In the event of a temporary disruption affecting Treasury data, resolution may be delayed until the official figures become available.