Moody's Talks - Inside Economics Weighing Recession Probabilities
Mar 27, 2026
Shandor Whitcher, an economist who built a random forest-based recession probability model, explains his approach and forecasts. He walks through which indicators matter most and recent model updates. The conversation covers market moves since the Iran conflict, the surprising role of the yield curve, and what would realistically tip the economy into recession.
AI Snips
Chapters
Books
Transcript
Episode notes
Geopolitical Shock Has Raised Recession Risks
- Recession probability has moved sharply amid the Iran war and market shocks.
- Mark Zandi and hosts note stocks down ~7–8%, 10-year yields up ~0.5 percentage points, and gasoline up ~$1 per gallon as accumulating damages.
Random Forest Model Distills Many Signals
- Shandor Whitcher built a random forest model to compress many indicators into a 12‑month recession probability.
- The model weights labor market most heavily, then financial and composite indicators, and also includes oil prices and housing.
Model Jumped After Jobs Then Settled Lower
- The model spiked to just over 48% after the February jobs release but later revised down to ~40% with more data.
- Shandor notes early runs after the jobs report can overstate risk because other monthly inputs are still missing.
