What is the model telling us right now?
The current stage summarizes how Growth, Labor, the yield curve and credit are lining up.
This commentary explains the model’s current classification without changing any rule or threshold.
The framework moves through six stages as economic stress builds. The current stage is highlighted below, followed by the full monthly history.
The current stage summarizes how Growth, Labor, the yield curve and credit are lining up.
This commentary explains the model’s current classification without changing any rule or threshold.
Select any month to inspect the signals behind that stage.
A month is missing from the underlying audit trail.
This is the model’s historical research view using the current final architecture and historical data available in the audit trail. Because economic data can be revised, this is not the same thing as the separate real-time/vintage replay used during validation.
This is the model’s main real-economy signal. Growth is measured primarily with Industrial Production, while Labor combines Unemployment and Initial Jobless Claims. The model does not escalate the regime because of one weak report — it looks for weakness to persist.
| Month | Stage | Growth | Labor | Persistent Growth + Labor Stress? | Curve | Credit | Sahm | Recession |
|---|---|---|---|---|---|---|---|---|
The model is a six-stage economic-risk framework. It is built to separate normal monthly noise from weakness that is becoming persistent and broad enough to matter.
Whether stress is spreading across real activity, labor and financial conditions — and whether those signals are confirming one another. The output runs from Expansion to Contraction; it is not a short-term stock-market signal.
Each input covers a different part of the cycle: Industrial Production for real activity; unemployment and jobless claims for labor deterioration; the 10Y–3M curve for prior monetary and financing pressure; Baa credit spreads for market confirmation; and the Sahm rule for late labor confirmation.
More indicators do not automatically make a better model. The model favors a small set of distinct signals, persistence tests and cross-confirmation so one noisy release cannot move the regime by itself.
Analytical framework only — not a recession declaration, probability forecast or investment recommendation.