10Y Treasury vs 10Y-3M Spread

The April 2026 readings: 10Y at 4.31 percent, 3M T-bill at 3.68 percent, T10Y3M spread at 63 basis points. The 3M yield sits within 7 basis points of the upper bound of the fed funds range (3.50 to 3.75 percent), reflecting normal money-market arbitrage. The T10Y3M spread of 63 bps is well above the January 16, 2026 reading of 51 bps, indicating modest steepening through Q1 2026.

The steepening is consistent with three drivers: the Iran war oil-driven reflation supporting the long end, modest fiscal-deficit term premium accumulation, and the absence of fresh recession signals that would compress the long end. The 10Y has held a relatively narrow range of 4.20 to 4.50 percent through April, while the 3M has been similarly narrow at 3.65 to 3.75 percent.

The T10Y3M spread (10-year Treasury yield minus 3-month T-bill yield) is the spread used by the NY Fed for its monthly recession-probability model published since the early 1990s. The model produces a 12-month-forward recession probability based on the spread's level and historical relationship to NBER-dated US recessions.

The spread has preceded each of the last eight US recessions with inversion (spread below zero) producing a recession within 6 to 18 months. The 2022 to 2024 inversion (T10Y3M went negative in October 2022, peaked at minus 188 bps in May 2023, returned to positive in December 2024) is the longest sustained inversion in the model's history. The fact that recession did not materialize during this inversion is the central recent debate in macro forecasting. Many factors have been proposed: COVID-related labor supply disruption, fiscal stimulus persistence, AI capex offsetting tightening, wealth-effect support from equity markets at all-time highs.

Pairing the 10Y level with T10Y3M creates a richer signal than either alone. Three regime combinations are diagnostic. First, high 10Y plus narrow positive T10Y3M (current configuration): suggests both fiscal-driven term premium and Fed policy near neutral, with no recession warning. Second, low 10Y plus deeply inverted T10Y3M (2007 to 2008, 2019 to 2020): suggests the market expects aggressive Fed cuts on recession risk, with growth expectations (10Y level) already declining. Third, low 10Y plus positive T10Y3M (2009 to 2014, 2020 to 2021): suggests a Fed easing cycle either underway or completed, with the curve restored to its normal shape and economy in recovery.

The current high-10Y plus positive-T10Y3M configuration is the textbook "expansion" reading: economy expanding, inflation contained but elevated, Fed cautious about cutting too aggressively, fiscal supply pressuring long end.

T10Y3M inverted in October 2022 with the Fed mid-hiking-cycle at fed funds of 3 to 3.25 percent and 10Y at 3.85 percent. The inversion deepened as Fed continued hiking through July 2023 (final hike to 5.25 to 5.50 percent), with T10Y3M peaking at minus 188 bps in May 2023.

The inversion held through 2023 and most of 2024. The Fed cut 100 bps from September to December 2024 (5.50-5.25% down to 4.50-4.25%), and continued cutting through 2025 (to 3.50 to 3.75 percent), bringing the front end down faster than the long end. T10Y3M returned to positive in December 2024 and has steepened steadily since. The current 63 bps spread reflects 18 months of Fed cuts (200 bps total) and stable long-end yields.

The NY Fed recession-probability model peaked at approximately 70 percent during the deepest inversion in May 2023. The model now reads approximately 5 to 8 percent recession probability over the next 12 months, the lowest since 2022.

The unprecedented 26-month inversion (October 2022 through December 2024) without recession is the central debate in yield-curve forecasting. Three frameworks for understanding this episode:

First, the false-signal interpretation: structural changes (low full-cycle Treasury issuance, abundant Fed reserves, inverted relationship between term premium and short rates) have broken the historical link between inversion and recession. The model that worked 1968 to 2020 may not work post-2022.

Second, the delayed-signal interpretation: the recession is still coming but the lag has extended due to fiscal stimulus, COVID savings drawdown, and AI capex masking underlying weakness. Watch for ISM rolling, payrolls turning negative, or sudden financial-conditions tightening to confirm the eventual recession.

Third, the "Goldilocks" interpretation: the Fed managed to engineer a soft landing, breaking the historical inversion-recession pattern. Inflation came down without forcing recession because supply chains normalized post-COVID and labor force participation recovered.

The April 2026 60-bp positive spread is consistent with frameworks 1 or 3. Framework 2 remains a tail risk that watches for ISM and payrolls rolling.

The NY Fed model uses 3-month T-bill rather than fed funds or 2Y Treasury for two reasons. First, 3M T-bill is the closest market-traded proxy for current monetary policy stance: it absorbs near-term Fed expectations (next 1 to 2 meetings) while filtering out longer-term cycle expectations. Second, the historical 3M data series is consistent and clean, while fed funds rates have changed how they are measured (effective fed funds vs target, vs implied fed funds, etc.) over decades.

The T10Y3M spread also has stronger empirical recession signal than T10Y2Y: it captures pure Fed-policy stance (3M is essentially fed funds plus a few basis points) versus market expectations of where rates will be in 2 years. Estrella and Mishkin (1996, 1998 NY Fed papers) demonstrated T10Y3M outperforms T10Y2Y in out-of-sample recession forecasting. The Cleveland Fed yield curve recession model uses T10Y3M.

From December 2024 (T10Y3M re-positive at approximately 0 bps) through April 2026 (63 bps), the spread has steepened by approximately 60 basis points. The steepening came almost entirely from the 3M side: 3M fell from approximately 4.30 percent (early Dec 2024) to 3.68 percent (April 2026), a 62 bp decline matching Fed cuts. The 10Y has been roughly flat at 4.30 to 4.40 percent over the same period.

The steepening is the classic post-easing-cycle pattern: front end falls with Fed cuts, long end stays anchored on fiscal and structural factors. The Cleveland Fed yield-curve recession-probability model now reads approximately 5 percent (down from 70 percent peak in mid-2023). The Conference Board Leading Economic Indicators have stabilized after declining 2022 to 2024. ISM Manufacturing PMI has held above 50 most of 2025 to 2026 (March 2026 reading 51.3).

The 10Y level matters independently of the curve shape. Three components determine the 10Y level: expected average fed funds over 10 years (the policy-path component), term premium (compensation for duration risk), and any liquidity premium specific to the 10Y benchmark.

In April 2026, the 10Y at 4.31 percent decomposes approximately as: expected fed funds path averaging 3.5 percent over 10 years (consistent with neutral fed funds estimates) plus 80 to 90 bps of term premium plus 0 bp liquidity premium (10Y is the most liquid Treasury point).

For context, the 10Y at 4.31 percent is approximately 100 bps above its 2010 to 2021 average of 2.30 percent and 30 bps below its 2007 pre-financial-crisis level of 4.65 percent. The current level is consistent with nominal GDP growth expectations of 3.5 to 4 percent (real growth 1.5 to 2 percent plus inflation 2 to 2.5 percent) plus a normal-sized term premium.

Five regimes summarize the historical pair behavior. Regime 1 (early-cycle expansion, 2010 to 2014): 10Y low 2 to 3 percent, T10Y3M wide 200 to 300 bps. Regime 2 (mid-cycle expansion, 2014 to 2018): 10Y rising 2 to 3 percent, T10Y3M narrowing 100 to 200 bps. Regime 3 (late-cycle, 2018 to 2019): 10Y peaking near 3.25 percent, T10Y3M briefly inverted in March 2019. Regime 4 (recession, 2020): 10Y collapsing to 0.5 percent, T10Y3M re-steepening. Regime 5 (post-recession recovery and hike cycle, 2021 to 2024): 10Y rising 1.5 to 5 percent, T10Y3M from 200+ bps to minus 188 bps inversion.

The current configuration (10Y 4.31 percent, T10Y3M positive 63 bps) is most similar to Regime 2 (mid-cycle expansion) plus elevated 10Y level reflecting fiscal-driven term premium. The 12-month forward recession probability from the NY Fed model is approximately 5 to 8 percent, the lowest since 2022.

Five drivers will shape both the 10Y level and the T10Y3M spread through year-end 2026. First, FOMC decisions: each Fed cut compresses 3M directly, steepening T10Y3M; the dot plot revisions update both 10Y and 3M expectations. Second, Treasury refunding announcements: tilts toward longer maturities raise 10Y term premium without affecting 3M. Third, fiscal trajectory through FY2027 budget release.

Fourth, the Iran war duration: resolution would compress 10Y by 15 to 25 bps without much 3M impact, narrowing T10Y3M to 35 to 45 bps. Fifth, recession indicators: ISM rolling below 50 with payrolls negative would prompt Fed-cut acceleration, compressing 3M faster than 10Y, steepening T10Y3M to 100+ bps as a recession-warning re-emerges.

The base case: 10Y holds 4.20 to 4.60 percent range, T10Y3M holds 50 to 80 bps range. Tail scenarios: 10Y above 4.80 percent on fiscal deterioration; T10Y3M re-inverting only if Fed re-accelerates hiking on inflation surge.