DGS10 vs Fed Funds Rate: Monetary Transmission and the Yield Curve Mechanism

Term structure theory says DGS10 is the integral of expected Fed Funds plus a term premium. The empirical 2000-2026 relationship shows an unstable elasticity, periods of decoupling, and the rare configuration where DGS10 falls below the policy rate — followed by a recession in four of five historical episodes.

Reading the DGS10/Fed Funds relationship means reading monetary transmission itself, in its successes and its failures.

1. The theoretical formula and what it hides

Term structure theory states that the yield on a T-year bond equals the average of short rates expected over that horizon, plus a term premium compensating the holder for rate variation risk. Applied to the 10-year, the identity reads: DGS10 = expected average of Fed Funds over 10 years + 10-year term premium. The formulation is mathematically correct but operationally misleading if the second term is neglected.

The 10-year term premium published daily by the Federal Reserve Bank of New York, via the Adrian-Crump-Moench (ACM) decomposition, spent most of 2010-2021 in negative territory, reaching -100 bps in March 2020. A negative term premium means Treasury holders accepted a yield below the expected Fed Funds average — a phenomenon tied to QE programs, post-2008 safe-asset demand, and Basel III prudential regulation that required banks to hold a High Quality Liquid Assets stock primarily composed of Treasuries.

Since 2022, the ACM term premium has returned to positive territory, reaching +60 bps by end-2023 and oscillating around +30 to +50 bps in 2025-2026. This transition radically alters DGS10/Fed Funds elasticity: an identical move in expected Fed Funds now produces different effects on DGS10 depending on whether the term premium is stable, compressing, or expanding. This context dependence is readable through DGS10 as the long-end financing anchor in the real economy.

2. The empirical DGS10/Fed Funds elasticity, 2000-2026

Measuring DGS10 beta to the Fed Funds Rate requires choosing a measurement window and a variation mode (level vs change, instantaneous vs smoothed). Over 3-year rolling windows in monthly variation, the DGS10/DFF beta between 2000 and 2026 fluctuates in a 0.3 to 1.2 range. This dispersion is not noise: it reflects distinct macroeconomic configurations.

The 2004-2006 cycle saw Greenspan speak of a “conundrum”: the Fed raised Fed Funds from 1.0% to 5.25%, but DGS10 only followed at 30-40% of the move. The ex-post explanation — structural demand from Asian central banks recycling trade surpluses into Treasuries — was fully understood only after 2008. The 2015-2018 cycle saw an intermediate beta around 0.5-0.6: the Fed moved from 0.25% to 2.50%, DGS10 from 1.7% to 3.2%. The atypical 2022-2023 cycle showed an initial beta near 0.8 on the upleg (Fed Funds from 0.25% to 5.5%, DGS10 from 1.8% to 4.8%), followed by partial decoupling when the Fed paused.

This beta instability reflects the interaction between three variables: Fed Funds expectations (captured by Fed Funds futures and SOFR futures), the term premium (captured by ACM), and the composition of Treasury demand (captured by TIC data and bid-to-cover ratios). A rigorous monetary transmission analysis integrates these three variables, not just the simple DGS10/DFF correlation.

This contextual dependence of the DGS10/DFF beta has a direct consequence for reading Fed decisions. A 25 bps Fed Funds hike in 2024 produces a different DGS10 effect than a 25 bps hike in 2007 — and the difference owes not only to expectations but to term premium resilience in each regime. This property makes DGS10 a synthetic indicator of the health of monetary transmission itself: when DGS10 fails to follow Fed Funds, it means either that markets contest the announced trajectory, or that the term premium absorbs the move.

3. The rare configuration of DGS10 below Fed Funds

In May 2026, DGS10 stands around 4.2% while the Fed Funds Rate target remains at 4.25-4.50%. This configuration, where the 10-year yield falls below the policy rate, is statistically rare. Across 64 years of series, it has been observed in only five distinct episodes: 1979-1980 (Volcker tightening), 1989 (pre-1990 recession), 2000 (pre-dot-com burst), 2006-2007 (pre-subprime crisis), and the 2022-2024 then 2024-2026 sequence.

Four of the five historical precedents were followed by a recession within 12 to 24 months — a historically high prediction rate for a market signal. But the statistic rests on five observations only, and each episode sits in a distinct macroeconomic context: anti-inflation fight 1979-1980, end of Greenspan cycle 1989, dot-com bubble 2000, housing imbalances 2007. The 2024-2026 configuration does not replicate any of these contexts identically, calling for caution in mechanical extrapolation. The complete signal analysis is treated in the 10-year minus 3-month recession signal.

The May 2026 mixed configuration — DGS10/DFF slope negative but T10Y3M slope re-positive since August 2024 — challenges the classical reading grid. The DGS10/Fed Funds relationship across three regimes historically shows that each episode has its own signature, and that the recession signal is most robust when both T10Y3M and DGS10/DFF are simultaneously negative (the 1979, 1989, 2000, 2007, 2022 configuration).

4. The decisive role of the term premium

The term premium is the conceptual bridge that reconciles theory and empirics. When stable, the DGS10/Fed Funds elasticity approaches theoretical values (close to 1 on long windows). When it moves, the elasticity collapses or amplifies. Episodes where the term premium rebuilds rapidly (2022-2023) coincide with observable decouplings between expected Fed Funds and the published DGS10.

The September-October 2023 episode illustrates this point. Fed Funds futures were already pricing aggressive cuts for 2024 (an implicit 100-125 bps reduction), suggesting that the “Fed Funds expectations” component of DGS10 should fall. Yet DGS10 rose from 3.8% to 5.0% over six weeks. The ACM decomposition showed that the term premium had rebuilt +50 bps over the same period, contributing more than half of the total move. The Fed explicitly cited this autonomous “long-end tightening” as contributing to its hike pause at the 1 November 2023 FOMC.

This substitution logic between Fed Funds and term premium is a monetary transmission mechanism in its own right. When the market reprices the term premium upward, the Fed can slow the pace of tightening. When the term premium compresses, the Fed may instead need to accelerate to maintain equivalent transmission. This is precisely the policy transmission channel to corporate balance sheets that links Fed Funds to corporate balance sheets via DGS10.

For fixed income portfolio holders, this dynamic has practical implications. The effective duration of the 10-year Treasury varies depending on whether the term premium is stable or volatile. During phases of rapid term premium rebuild (2022-2023, October 2023), observed 10-year convexity is higher than in a stable regime, modifying the real risk-return profile of duration strategies. Treasury demand pressure on the term premium details how fiscal supply interacts with these dynamics.

5. Why this relationship differs from classical inversion

It is essential to distinguish the DGS10/Fed Funds relationship analyzed here from the yield curve inversion reading. Classical inversion compares DGS10 to DGS3M (the 3-month CMT yield), not to the effective Fed Funds Rate (DFF). The two are linked — DGS3M closely tracks DFF — but the gap between them can be substantial during money market stress episodes.

When the Fed conducts its open market operations via the Reverse Repo Facility and the discount window, DGS3M can diverge by 5-15 bps from the effective Fed Funds rate. This gap is generally small but can reach 30-50 bps during Bill market tensions (March 2020, September 2019 on repo). For monetary transmission analysis, the DGS10/DFF relationship isolates the impact of Fed decisions on the Treasury curve; the DGS10/DGS3M relationship adds money market noise.

This distinction has implications for reading the recession signal. T10Y3M, the academic reference signal (NY Fed, Cleveland Fed, Conference Board), uses DGS3M rather than DFF. Its statistical robustness across multiple cycles comes partly from the fact that DGS3M already incorporates very-short-horizon expectations — the Bill market prices expected Fed Funds over 3 months. The DGS10/DFF spread is more immediate but less forward-looking than T10Y3M.

A practical implication of this distinction is the timing of recession signals. T10Y3M tends to invert several quarters before DGS10/DFF, because the 3-month Bill yield reacts faster to expected Fed pauses or cuts than the effective Fed Funds rate which only moves when the FOMC formally acts. Tracking both signals therefore provides a leading-lagging pair that helps identify the maturity of the cycle: T10Y3M inverts first (forward expectation), DGS10/DFF inverts later (current policy), and re-steepening of T10Y3M typically precedes Fed Funds cuts by 6 to 12 months in historical episodes.

Last updated — 19 May 2026