Recession Probability vs Unemployment
NY Fed Recession Probability (12-months-ahead, derived from 10Y-3M Treasury spread) measures forward recession risk from yield curve dynamics. Unemployment Rate (U3, FRED UNRATE) measures realized labor market slack.
Also known as: CVRP, Convex Recession Probability (CVRP, CRPI, recession probability, recession index) · Unemployment Rate (U3) (unemployment, U3, jobless rate)
Why This Comparison Matters
NY Fed Recession Probability (12-months-ahead, derived from 10Y-3M Treasury spread) measures forward recession risk from yield curve dynamics. Unemployment Rate (U3, FRED UNRATE) measures realized labor market slack. April 2026: NY Fed recession probability approximately 18.8 percent (down from 20.7 percent February 2026 reading; declining trend). Unemployment rate approximately 4.3 percent (Sahm Rule triggered July 2024 but no recession 21+ months later). Recession probability is forward (yield curve embeds expectations); unemployment is contemporaneous-to-lagged. Combined: recession risk moderate (~19 percent) + unemployment moderately elevated. Yield curve has re-steepened from 2022-2023 inversion lowering forward recession probability. Late-cycle but not recessionary.
The April 2026 Configuration
NY Fed Recession Probability: ~18.8% (April 4, 2026, using March 2026 data). Down from 20.7% February 2026. Declining trend reflects yield curve re-steepening (10Y-3M spread positive again).
Unemployment rate (U3): ~4.3% (April 2026). Up from 3.5% September 2022 cycle low. Sahm Rule triggered July 2024 (3-month MA 0.5pp above 12-month MA low). Has been triggered for 21+ months without recession (longest period without recession after Sahm trigger in 54-year history).
10Y-3M Treasury spread (NY Fed model input): currently positive ~70bp (10Y 4.31% minus 3M ~3.55%). Compares to peak inversion -150bp+ in 2023.
The combined April 2026 reading: forward recession probability moderate (19%) + unemployment elevated but stable (4.3%). Late-cycle expansion. Sahm Rule false positive territory. Recession not imminent per yield curve model but elevated risk.
Long-Term Range and Recent Trajectory
NY Fed recession probability range: 0% to 70% historical. Above 30% historically preceded recessions (12-month lead). Above 50% near-certain. Below 10% economic expansion.
2022-2023 spike: peaked at 70%+ June 2023 reflecting -150bp 10Y-3M inversion. Highest since 1981. Historical model would have signaled recession Q3 2024-Q1 2025. Did not arrive.
2024-2026 normalization: declined from 70%+ to 18.8% currently. Reflects yield curve re-steepening + persistent economic resilience. NY Fed model probability has fallen ~50 percentage points without recession arriving.
Unemployment trajectory: 3.5% (September 2022 low) to 4.3% (April 2026). +80bp rise over 3.5 years. Slow grinding higher. Labor force expansion 3-4M workers absorbed without recession.
Sahm Rule history: triggered July 2024 (3-month MA U3 0.5pp above prior 12-month low). Historically 100% accurate recession signal. April 2026: 21+ months since trigger without recession. Most divergent Sahm Rule reading in history.
Historical Precedents: Past Episodes
2008-09 GFC: NY Fed probability reached 40%+ peak December 2007. Recession started December 2007 (NBER). Unemployment 4.6% to 10.0% (2009). Yield curve inversion peaked late 2006 at -50bp.
2020 COVID: NY Fed probability ~30% before COVID shock. Unemployment 3.5% to 14.7% (April 2020). Recession Q1-Q2 2020. Sudden shock not predicted by yield curve.
2001 dot-com: NY Fed probability peaked 50% August 2000. Recession March-November 2001. Unemployment 3.9% to 6.3%. Yield curve inverted -75bp peak.
1990-91: NY Fed probability peaked 35% mid-1989. Recession July 1990-March 1991. Unemployment 5.0% to 7.8%.
2022-2026 anomaly: NY Fed probability peaked 70%+ June 2023 (highest since 1981). Sahm Rule triggered July 2024. No recession 21+ months later. Unprecedented divergence between yield curve recession signals + economic resilience. Reasons: services-driven economy, immigration labor expansion 3-4M workers, AI capex ~$300B+ annual, fiscal support, Fed easing room.
Mechanics: Why Yield Curve Predicts Recessions
Yield curve mechanics: short rates set by Fed reflect current monetary stance. Long rates reflect expected future short rates + term premium + inflation expectations. Inversion (long below short) implies market expects Fed cuts = economic weakness expected.
NY Fed 12-month model: regression of recession indicator on 10Y-3M spread historical 1959-present. Inversion below zero = elevated probability. -150bp inversion = ~70% probability.
Lag structure: yield curve inverts ~12-18 months before recession historically. NY Fed model uses 12-month lookahead.
Unemployment lag: recessions cause unemployment to rise + unemployment continues rising 6-12 months after recession ends. Lagging indicator. Sahm Rule (3-month MA 0.5pp above 12-month low) triggers near recession start typically.
2022-2023 yield curve inversion may be most clear false signal in 50+ years. Reasons: (1) Fed hiked aggressively (525bp in 14 months) creating inversion mechanically, (2) services economy less rate-sensitive, (3) fiscal/AI capex offset rate drag, (4) labor force expansion absorbed slack.
Reading the Pair: Convergence and Divergence
Convergence regime: low recession probability + low unemployment = early-mid cycle expansion. Best risk-on. Examples: 2010-2014, 2017-2019.
Divergence type 1: high recession probability + low unemployment = forward-looking model warning before labor market deterioration. Historical pattern 6-12 months before recession. Examples: late 2007, mid-2000.
Divergence type 2: low recession probability + rising unemployment = post-recession recovery. Yield curve normalized, labor market still healing. Examples: 2009-2010, late 2020.
Divergence type 3: high recession probability + rising unemployment but no recession = current April 2026 anomaly OR pre-recession setup. Difficult to distinguish in real time.
April 2026 regime: NY Fed probability declined 70%+ to 18.8% + Sahm triggered but unemployment stable + no recession. Most divergent setup in 54-year history. Either soft landing confirmed OR delayed recession arrives unexpectedly.
Driver Decomposition: What Moves the Spread
Driver 1: Fed policy stance. April 2026 Fed funds 3.50-3.75% (paused since December 2024). Down from 5.25-5.50% peak. Easing supports yield curve steepening + lower recession probability.
Driver 2: Long-rate dynamics. 10Y 4.31%. Sticky high reflecting fiscal trajectory + term premium + inflation expectations. 10Y above pre-COVID 2.0% average.
Driver 3: Labor force expansion. 3-4M workers added 2022-2026 (immigration + post-pandemic). Absorbed labor demand. Pushed unemployment higher without recession.
Driver 4: Productivity. AI-driven productivity gains supporting non-inflationary growth. GDPNow Q1 2026 +1.2%.
Driver 5: Fiscal. ~$2T deficits sustained. Supportive of demand. Offsets monetary tightness.
Driver 6: AI capex. ~$300B+ annual investment. Direct GDP contribution. Hyperscalers Apple Nvidia Microsoft Google Meta Amazon driving capex boom.
Net: yield curve normalization (lower recession probability) + persistent labor expansion (higher unemployment without recession) = late-cycle but not recessionary regime.
Cross-Asset Implications
Bonds: 10Y 4.31% reflects sticky yields despite Fed pause. Bond market not pricing aggressive Fed cuts. Curve re-steepening provides positive carry.
Dollar: DXY ~100. Mild dollar strength reflecting US growth differential. Recession probability decline supportive of dollar.
Equities: SPY ~$712 record territory. Equities priced for soft landing. Recession probability decline equities-supportive.
Commodities: Gold $4,722 record. Reflects monetary debasement + geopolitical hedge demand more than recession hedge.
Volatility: VIX 18.76 (elevated but not stressed). Markets pricing moderate risk.
Credit: HY spreads tight. Inconsistent with recession risk.
April 2026 cross-asset reading: all asset classes positioned soft-landing. Recession probability decline (NY Fed 18.8%) consistent with cross-asset positioning. Sahm Rule trigger appears to be false signal in this cycle.
Trading the Pair: Setups and Sizing
Setup 1 (soft landing confirmed, base case 65%): NY Fed probability declines to <15%, unemployment stabilizes 4.0-4.5%, no recession. Trade: long SPY + long cyclicals (XLI, XLF) + flatten yield curve carry. Profit from continued expansion.
Setup 2 (delayed recession arrives, risk 25%): unemployment breaks above 5%, NY Fed probability re-accelerates above 30%, initial claims above 350K. Trade: short SPY + long bonds (TLT) + long volatility. Aggressive recession positioning.
Setup 3 (status quo divergence, 10%): recession probability stays elevated but unemployment stable. Trade: balanced positioning, watch for resolution.
Key watch points: NY Fed monthly probability release (1st week each month), unemployment monthly (1st Friday), yield curve daily, initial claims weekly.
Position sizing: in late-cycle uncertainty, reduce gross exposure 10-15% from neutral. Keep duration neutral. Reserve for setup 2 confirmation.
Convex Indices Linkage
Convex Recession Probability Index (CVRP): synthesizes NY Fed yield curve probability + Sahm Rule + LEI + claims momentum + credit spreads. April 2026 CVRP elevated (Sahm Rule triggered, LEI declining, NY Fed probability moderate).
Convex Net Liquidity Impulse (CNLI): Fed balance sheet + reverse repo + TGA. April 2026 CNLI neutral-positive (Fed paused, RRP drained, fiscal supportive). Tailwind to growth.
Convex Risk Appetite Index (CRAI): credit spreads + equity volatility + risk currencies. April 2026 CRAI elevated (HY tight, VIX 18.76, AUD/JPY firm). Risk-on positioning.
Divergence between CVRP (recession concerns) + CRAI (risk-on) characterizes late-cycle. Use CVRP as early-warning. CRAI for trade timing.
April 2026 reading: late-cycle complacency or genuine soft landing. Cross-asset markets discount soft landing. Watch for CVRP escalation as recession trigger.
What to Watch in 2026
NY Fed probability trajectory: declining trend continues = soft landing. Re-acceleration above 30% = recession warning.
Unemployment direction: stable 4.0-4.5% = healthy expansion. Above 5.0% = labor market crack.
Initial claims: above 250K = weakening. Above 350K = recession imminent. April 2026 ~225K stable.
Fed cuts: market pricing 1-2 cuts H2 2026. If 75bp+ arrive, yield curve steepens further, recession probability declines, soft landing confirmed.
Yield curve: 10Y-3M spread above 50bp positive = soft landing. Below zero re-inversion = recession warning re-engaged.
Sahm Rule: continued trigger without recession = unprecedented anomaly continues. Resolution either: (1) Sahm declared false signal in this cycle OR (2) delayed recession finally arrives.
Geopolitical/exogenous: Iran tensions, China-Taiwan, oil shocks could change forward indicators rapidly.
April 2026 base case: recession probability declines to <15%, unemployment stabilizes 4.0-4.3%, soft landing confirmed by year-end.
90-Day Statistics
Explore Each Metric
Related Scenarios & Forecasts
Get daily macro analysis comparing key metrics delivered to your inbox. Stay ahead of market-moving divergences.
Frequently Asked Questions
What is the April 2026 NY Fed recession probability vs unemployment configuration?+
NY Fed recession probability ~18.8% (April 2026, declining from 20.7% February). Unemployment ~4.3% (up from 3.5% September 2022 low). Sahm Rule triggered July 2024, has been triggered 21+ months without recession (longest in 54-year history). 10Y-3M spread positive ~70bp (recovered from -150bp peak inversion 2023). Late-cycle expansion. Soft-landing trajectory.
How does the NY Fed model work?+
NY Fed Recession Probability uses 10Y-3M Treasury spread to predict recession 12 months ahead. Regression of recession indicator on yield curve spread historical 1959-present. Inversion (long below short) implies expected Fed cuts = weakness expected. -150bp = ~70% probability. Currently +70bp = 18.8% probability. Released monthly first week.
Why has Sahm Rule trigger not produced recession?+
21+ months since July 2024 trigger without recession is unprecedented. 1948-2024: every Sahm trigger preceded recession. April 2026: longest divergence in 54-year history. Reasons: (1) labor force expansion 3-4M workers (immigration + post-pandemic) raised U3 without job losses, (2) services-driven economy less recession-prone, (3) AI capex ~$300B+ annual + fiscal support sustained demand, (4) Fed easing room from 5.25-5.50% peak.
How do recessions historically progress in unemployment?+
2008-09 GFC: U3 4.6% to 10.0% peak. +540bp rise. 2020 COVID: U3 3.5% to 14.7% peak (sudden shock). 2001 dot-com: U3 3.9% to 6.3%. +240bp. 1990-91: U3 5.0% to 7.8%. +280bp. April 2026: U3 3.5% to 4.3% +80bp. Slow grinding rise without recession-style spike. Pattern more consistent with healthy adjustment than recession.
What does the divergence imply for positioning?+
Late-cycle divergence warrants moderate caution. Reduce gross exposure 10-15% from neutral. Keep bond duration neutral. Watch for resolution. Setup 1 (base case 65%): soft landing confirmed, NY Fed probability falls below 15%, U3 stable. Long equities + cyclicals. Setup 2 (risk 25%): delayed recession arrives, U3 above 5%, initial claims above 350K. Short equities + long bonds + long volatility. Setup 3 (status quo 10%): divergence persists.
Is the 2022-2023 yield curve inversion a false signal?+
Most clear false signal in 50+ years. -150bp peak inversion June 2023 (highest since 1981) historically signals recession with near-certainty. April 2026: 32 months later no recession. Reasons: (1) Fed hiked aggressively 525bp creating mechanical inversion not reflecting fundamental weakness, (2) services economy less rate-sensitive, (3) fiscal + AI capex offset rate drag. Yield curve recession signal may need recalibration for post-COVID economy.
How is the pair used for trading?+
NY Fed probability declining + U3 stable: soft landing confirmed. Long equities, long cyclicals, flatten yield curve carry. NY Fed probability rising + U3 stable: late-cycle warning. Reduce gross. NY Fed probability rising + U3 rising: recession setup. Short equities, long bonds, long volatility. NY Fed probability declining + U3 rising: post-recession recovery. April 2026 regime: probability declining + U3 stable = base case soft-landing positioning.
What macroeconomic conditions explain the 2022-2026 anomaly?+
Five structural changes weakened traditional recession signals: (1) services-driven economy (~70% GDP) less manufacturing-cycle dependent, (2) immigration + post-pandemic labor force expansion 3-4M workers absorbed slack, (3) AI capex ~$300B+ annual sustained demand, (4) fiscal deficits ~$2T continuing supportive, (5) Fed easing room from 5.25-5.50% peak to 3.50-3.75%. Combined: traditional yield curve + Sahm Rule recession predictors are giving false positives in this cycle.
Related Comparisons
Explore Across Convex
Data sourced from FRED, CoinGecko, CBOE, and other providers. This page is for informational purposes only and does not constitute financial advice. Past performance does not guarantee future results.