Methodology and Sources

• US recession definitions (NBER and the popular two-quarter rule) and the historical record of where they diverge.
• Real-time US recession indicators: Sahm rule, yield curve, jobless claims, ISM PMI, LEI, consumer confidence, unemployment trend, HY credit spreads, housing starts.
• Post-WWII duration and depth data for each of the 13 post-1945 NBER-dated US recessions, with peak unemployment and recovery duration.
• Complete 34-recession NBER history from the 1857 Panic through the 2020 COVID recession.
• Personal US recession preparation: emergency fund sizing, debt prioritisation, HYSA cash management, job-skill investment, retirement-contribution discipline.
• Current 2026 outlook synthesis citing institutional model probabilities (NY Fed, Cleveland Fed) and economist surveys (Bloomberg, Goldman Sachs Research).

• Non-US recessions (UK, eurozone, China, Japan, emerging markets) beyond IMF / OECD framing references.
• Individual stock recommendations or named-security trade ideas.
• Market-timing advice (entry / exit calls, dollar-cost averaging schedules tied to indicator readings).
• Predictions of recession start dates beyond reporting institutional model 12-month probability outputs.
• Monetary-policy advocacy (positions on whether the Fed should cut, hold, or hike).
• Legal, tax, or fiduciary advice tied to personal recession-preparation decisions.

The Business Cycle Dating Committee weighs real GDP, real personal income less transfers, nonfarm payroll employment, household employment, real manufacturing-trade sales, and industrial production. Depth, diffusion across sectors, and duration are all considered. No single quantitative rule triggers a declaration; the committee exercises qualitative judgment over the indicator set.

The Sahm rule fires when the three-month moving average of the national unemployment rate (BLS CPS series UNRATE) rises 0.50 percentage points or more above its lowest reading from the prior 12 months. The current reading is published on FRED as SAHMCURRENT. As of April 2026 the reading is 0.47, just below the trigger.

Two consecutive quarters of negative real GDP growth is a widely cited proxy but not the official US definition. It produces the right answer roughly 80% of the time; the divergence cases are 2001 (NBER recession, no two-quarter trigger), 2020 (NBER recession lasted two months, GDP rule technically triggered but the recession ran inside one quarter), and 2022 H1 (two-quarter trigger fired, NBER declined to declare). The site reports both with the divergence cases named.

Recessions are grouped by peak-to-trough real GDP contraction: mild (below -2%), moderate (-2% to -4%), severe (-4% to -8%), and depression-level (more than -10%). The bands are central tendencies used for editorial framing, not predictions. Pre-WWII recessions are reported separately because the GDP measurement methodology and the institutional shock-absorbers (FDIC, automatic stabilisers) differ materially.

Recovery duration is measured peak-to-prior-peak (the time from the pre-recession GDP peak back to that GDP level), not peak-to-trough. Employment recovery duration is reported separately (peak nonfarm payrolls back to that payroll level), and is consistently longer than GDP recovery duration in financial-crisis recessions.

The /calculator self-assessment scores eight weighted inputs (emergency fund months, debt-to-income ratio, job tenure, sector resilience, retirement contribution discipline, portfolio diversification, written budget, income diversification). Sector resilience is calibrated against BLS-reported layoff rates by NAICS industry across the 2007-09 and 2020 recessions. The calculator is a self-assessment, not a probability model.