Regime-Dependent Recession Forecasts with Six-Month-Ahead Signals

Michael J. Dueker of the Federal Reserve Bank of St. Louis developed a six-month-ahead recession forecasting signal designed to reduce false alarms. The article evaluates this signal's out-of-sample performance, including a prediction close to the March 2001 recession onset. The dataset likely contains the time-series indicators and signals used for this analysis.

Use Cases

• Evaluating recession forecasting models based on the described six-month-ahead signal.
• Comparing the performance of different economic indicators for predicting the 1990-1991 and 2001 recessions.
• Analyzing the trade-off between false signals and true recession detection in economic models.
• Studying regime-dependent forecasting techniques mentioned in the article's title.

Strengths

• Method developed by an author affiliated with the Federal Reserve Bank of St. Louis, suggesting institutional expertise.
• Specific out-of-sample evaluation against the 1990-1991 and 2001 recessions is documented.
• The signal is designed with a stated goal of reducing false alarms outside of recessions.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and dataset size are unknown, which may limit suitability assessment.
• Last update date is unknown; freshness unverified.

Provenance

Source

Michael J. Dueker, Federal Reserve Bank of St. Louis

Collection Method

Likely derived from published economic indicators and time-series analysis, as described in the associated article.

Time Range

Analysis period includes the 1990-1991 and 2001 recessions.