Recession Prediction: Leading Indicators & Early Warning Signals

The Recession Prediction Challenge

Predicting recessions is hard:

• Officially declared months after they start
• Every recession is different
• False positives costly
• Missing it = worse

But we can improve odds with data.

Official Recession Dating

NBER Business Cycle Dating Committee

What defines recession:

"Significant decline in economic activity spread across the economy, lasting more than a few months."

Key indicators NBER uses:

• Real personal income less transfers
• Nonfarm payroll employment
• Real personal consumption expenditures
• Real manufacturing and trade sales
• Industrial production
• Aggregate hours worked

The Lag Problem

NBER declares recessions ~6-12 months after they start.

Markets can't wait—need real-time signals.

Classic Leading Indicators

Yield Curve Inversion

FRED Series: T10Y2Y (10Y minus 2Y) | IQ Score: 97

The most famous recession predictor.

The signal: When short-term rates exceed long-term rates.

Track record: Predicted every recession since 1970.

Lead time: 6-24 months (variable).

Key spreads to watch:

• 10Y - 2Y (popular)
• 10Y - 3M (Fed prefers)
• 10Y - Fed Funds

Current interpretation:

Initial Jobless Claims

FRED Series: ICSA | IQ Score: 97

Weekly unemployment insurance claims.

The signal: Sustained rise from trough.

Rule of thumb: 50% rise from low = recession likely.

Lead time: Shorter than yield curve—months, not quarters.

ISM Manufacturing PMI

FRED Series: MANEMP (proxy) | IQ Score: 95

Key levels:

Conference Board Leading Economic Index (LEI)

FRED Series: USSLIND | IQ Score: 95

Composite of 10 leading indicators:

• Average weekly hours (manufacturing)

• Initial jobless claims

• New orders (consumer goods)

• ISM new orders

• New orders (capital goods)

• Building permits

• S&P 500

• Leading Credit Index

• Interest rate spread

• Consumer expectations

Three consecutive monthly declines = Warning signal.

Other Leading Indicators

Credit Conditions

SLOOS Survey (Senior Loan Officer Opinion Survey):

Net % of banks tightening standards.

Tightening precedes every recession.

Credit spreads:

• IG spreads
• High yield spreads
• Widening = stress

Labor Market

Sahm Rule:

FRED Series: SAHMREALTIME | IQ Score: 95

Recession starts when unemployment rate rises 0.5pp above its low of the previous 12 months.

Real-time, minimal false positives.

Consumer Indicators

Consumer confidence:

Sharp declines precede recessions.

Retail sales:

Real retail sales turning negative.

Housing

Housing starts:

Peak ~12 months before recession.

Building permits:

Lead starts by 1-2 months.

Composite Recession Probability Models

NY Fed Recession Probability

FRED Series: RECPROUSM156N | IQ Score: 94

Based on yield curve spread.

Probability of recession in next 12 months.

Interpretation:

Smoothed Recession Probabilities

FRED Series: USRECPROB | IQ Score: 93

Based on multiple indicators.

More stable than single-indicator models.

Private Sector Models

• Bloomberg Economics
• Goldman Sachs
• Morgan Stanley
• Academic models

Building a Recession Dashboard

Tier 1: High Priority (Weekly/Monthly)

• Yield curve (10Y-2Y, 10Y-3M)

• Initial jobless claims (weekly)

• LEI (monthly)

• NY Fed probability (monthly)

Tier 2: Confirming Indicators

• ISM Manufacturing (monthly)

• SLOOS (quarterly)

• Credit spreads (daily)

• Housing starts (monthly)

Tier 3: Coincident Confirmation

• Nonfarm payrolls (monthly)

• Industrial production (monthly)

• Real retail sales (monthly)

• Sahm Rule (monthly)

Interpretation Framework

Early Warning Phase

• Yield curve inverts
• LEI declining
• NY Fed probability rising
• No labor market weakness yet

Confirmation Phase

• Jobless claims rising
• Credit spreads widening
• Manufacturing contracting
• Consumer confidence falling

Recession Onset

• Payrolls declining
• Sahm Rule triggered
• Multiple coincident indicators weak
• (NBER will confirm later)

False Signals

1966-67 "Phantom Recession"

Yield curve inverted, no recession.

But manufacturing did contract.

1998 Near-Miss

Curve inverted, LTCM crisis.

Fed cut, avoided recession.

2019 Inversion

Curve inverted in August 2019.

COVID recession March 2020—coincidence?

Why False Positives Happen

• Policy response (Fed cuts)
• External shocks (positive)
• Structural changes
• International factors

Sector Implications

What Leads in Recessions

Outperformers:

• Treasuries
• Defensive equities (utilities, staples)
• Gold (sometimes)

Underperformers:

• High yield credit
• Cyclical equities
• Commodities (ex-gold)

Rotation Timing

Markets price recession before it starts.

Early movers have advantage.

Pro Tips

• No single indicator perfect: Use multiple signals

• Lead times vary: 6-24 months is wide range

• False positives hurt: Don't cry wolf

• Context matters: Each cycle different

• Policy response: Fed can extend cycle

• Global matters: External shocks can change picture

Explore recession indicators →