10Y-3M Spread vs 10Y-2Y Spread

Three readings define the April 2026 setup. T10Y3M at 63 bps: 10Y minus 3M, the NY Fed recession-model spread. T10Y2Y at 53 bps: 10Y minus 2Y, the market-favorite recession-signal spread. The 10 bp gap (T10Y3M minus T10Y2Y equals 63 minus 53) reflects 2Y at 3.78 percent sitting 10 bps above 3M at 3.68 percent.

The configuration is consistent with late-easing-cycle pricing. The 3M has fallen to 3.68 percent matching the 200 bps of cumulative Fed cuts since September 2024. The 2Y at 3.78 percent prices in essentially flat fed funds over the next 24 months: market view is that the Fed pause is durable but there is modest risk of another cut over 12-24 months. The 10 bp positive gap reflects that small remaining cut probability.

T10Y2Y reflects expectations of Fed policy 0 to 24 months out, with 2Y essentially equal to the average expected fed funds rate over the next 24 months. T10Y3M reflects current policy stance: 3M T-bill is essentially fed funds plus a few bps of liquidity premium.

The two spreads diverge during regime transitions. In hiking cycles, T10Y2Y inverts before T10Y3M because 2Y prices in eventual rate cuts before the Fed actually starts cutting. The 2022 to 2024 episode showed this clearly: T10Y2Y inverted in July 2022, T10Y3M inverted three months later in October 2022. In easing cycles, T10Y3M typically becomes steeper than T10Y2Y because 3M falls faster than 2Y as cuts deliver. The current 10 bp gap reflects exactly this dynamic.

Both spreads were deeply inverted simultaneously through 2022 to 2024, but at different magnitudes. T10Y2Y peaked at minus 109 bps in March 2023. T10Y3M peaked at minus 188 bps in May 2023, almost 80 bps deeper than T10Y2Y at the most extreme inversion.

The gap between them widened to roughly 80 bps during peak inversion, with T10Y3M deeper than T10Y2Y. This was the "Fed reaching peak rates" signal: 2Y was already pricing in eventual cuts (averaging fed funds across the next 24 months including expected cuts), while 3M tracked the actual peak fed funds at 5.25-5.50 percent. The 80 bp gap reflected approximately 160 bps of expected cuts over 24 months priced into 2Y.

The gap compressed through 2024 as Fed actually delivered cuts. By December 2024 both spreads had returned to approximately zero. Through 2025-2026 the gap has flipped: T10Y3M now steeper than T10Y2Y, reflecting easing-cycle dynamics.

In all eight recessions of the last fifty years, T10Y2Y has inverted before T10Y3M during the preceding hiking cycle. The lead time has averaged 3 to 6 months. Here is why: 2Y is forward-looking and prices in expected Fed cuts before they happen; 3M is backward-looking (tracks current fed funds).

This means T10Y2Y is the earlier-warning recession indicator, while T10Y3M is the more reliable confirmation. T10Y2Y produces more false signals (it can invert from temporary expectations of cuts that never come), while T10Y3M only inverts when the Fed has actually pushed fed funds well above the 10Y, which happens only during late hiking cycles.

For recession forecasting, the cleanest signal is T10Y3M sustained inversion (more than 30 days), which has 100 percent recession success rate over 8 episodes since 1968 (zero false positives, 8 successful predictions, with 6 to 18 month forecast horizon). Until the 2022-2024 episode, that is, which produced 26 months of T10Y3M inversion without recession.

The 2022 to 2024 episode is unique in T10Y3M history: 26 months of sustained inversion with no NBER-dated recession. This is the longest non-recessionary inversion in the series. T10Y2Y inverted simultaneously and produced the same false signal.

Three explanations dominate macro forecasting circles. First, structural breaks: post-2022 fiscal stimulus and AI capex cycle masked underlying weakness; the recession is delayed but still coming. Second, regime change: COVID-related labor supply disruption, abundant Fed reserves, and changed Treasury issuance composition broke the historical inversion-recession link. Third, soft landing: the Fed achieved Goldilocks for the first time since 1995 (the only previous soft landing).

The debate has practical significance for trading the curve. If interpretation 1 is correct, the current positive curve is misleading and recession is still likely 2026-2027. If 2 or 3 is correct, the positive curve accurately signals expansion and the model needs recalibration.

The T10Y3M-minus-T10Y2Y gap (T10Y3M minus T10Y2Y) tracks the position in the rate cycle remarkably reliably. Negative gap (T10Y3M deeper than T10Y2Y, currently positive 10 bps): late hiking cycle, peak rates approaching. Near-zero gap: hiking cycle complete or stable rates. Positive gap (T10Y3M steeper than T10Y2Y): easing cycle in progress, Fed cuts ongoing.

The 2022 to 2024 cycle showed this clearly. Mid-2022: gap approximately 0 bps (Fed early in hiking cycle). Mid-2023: gap minus 80 bps (T10Y3M deeper, peak hiking expectations). Late 2024: gap approximately 0 bps (Fed delivering cuts). April 2026: gap positive 10 bps (Fed pause with modest further-cut probability).

For curve traders, the gap moving from negative to zero is the "Fed has stopped hiking" signal. The gap moving from zero to positive is the "Fed is cutting" signal. The gap returning to zero from positive is the "easing cycle complete" signal.

Both spreads contain shared 10-year term premium components but differ in the front-end leg. T10Y2Y equals 10Y term premium plus expected fed funds at 10Y minus expected fed funds at 2Y. T10Y3M equals 10Y term premium plus expected fed funds at 10Y minus current fed funds approximately.

The difference T10Y3M minus T10Y2Y equals expected fed funds at 2Y minus current fed funds. This isolates the market's expected change in fed funds from now to 24 months from now. Currently, that expected change is approximately positive 10 bps, indicating markets expect fed funds slightly above current levels in 24 months (reflecting either modest hike expectations or stable rates with small term premium creep at the 2Y point).

This decomposition makes T10Y3M-minus-T10Y2Y one of the cleanest gauges of expected near-term Fed action, separated from term premium and longer-cycle expectations.

The Cleveland Fed yield-curve recession-probability model uses T10Y3M as its primary input. The Conference Board Leading Economic Indicators uses T10Y3M as its yield-curve component. The Atlanta Fed GDPNow nowcasting model does not use yield-curve directly but indirectly through financial-conditions indices.

T10Y2Y is more popular in financial media and trader-focused commentary because of its longer historical track record and tighter range (less volatile than T10Y3M). The 2-year futures and CME pricing make T10Y2Y trades more liquid and tighter spreads.

For a complete recession-signal read, watching both is essential. Both inverted (2022-2024) is the high-confidence recession warning; both positive (current) is the low-recession-probability all-clear; one inverted and one positive (e.g., March 2019 when T10Y3M briefly went negative while T10Y2Y stayed positive 15 bps) is the regime-transition warning that should trigger careful additional analysis.

The February 2026 Iran war has produced asymmetric effects on the two spreads. T10Y2Y narrowed approximately 5 bps from January 2026 (58 bps) to April 2026 (53 bps) as 2Y rose 25 bps on oil-driven inflation expectations while 10Y rose only 20 bps. T10Y3M widened approximately 12 bps from January 2026 (51 bps) to April 2026 (63 bps) as 3M was anchored to fed funds (no movement) while 10Y rose 20 bps.

The gap T10Y3M minus T10Y2Y went from approximately negative 7 bps in January 2026 to positive 10 bps in April 2026, a 17 bp swing. This reflects the Iran war making markets price in a shorter Fed pause (reducing the probability of additional cuts before 2027), which lifts 2Y closer to fed funds level.

Five drivers will shape both spreads through year-end 2026. First, FOMC dot-plot revisions: each meeting updates expected-fed-funds-2Y projections, moving 2Y by 5-15 bps. Second, NFP and CPI data: surprises move 2Y faster than 3M. Third, Iran war duration: resolution would compress 2Y by 15-20 bps (Fed cuts pricing back in), narrowing the T10Y3M-T10Y2Y gap. Fourth, recession indicators (ISM, payrolls, claims): rolling weakness would widen T10Y2Y faster than T10Y3M as 2Y falls on Fed-cut anticipation.

Fifth, Fed communication: hawkish surprises lift 2Y disproportionately to 3M; dovish surprises lower 2Y disproportionately. The base case: T10Y3M holds 50-80 bps, T10Y2Y holds 40-70 bps, with the T10Y3M-T10Y2Y gap oscillating between 0 and positive 20 bps. Tail scenario: gap flips negative to minus 20 bps if Fed signals re-acceleration of hiking on persistent inflation.