Multi-Horizon Forecast Comparison Tests for Macroeconomic Variables

Rogier Quaedvlieg introduces statistical tests for multi-horizon superior predictive ability (SPA). The article defines uniform and average SPA concepts and extends the model confidence set for comparing more than two models. An illustration on a large set of macroeconomic variables demonstrates the empirical benefits of multi-horizon comparison.

Use Cases

• Comparing forecasting model performance across multiple horizons based on the introduced uniform SPA concept.
• Evaluating average forecast path performance using the average SPA concept, which allows compensation across horizons.
• Applying the extended model confidence set to select the best-performing model from a set of more than two candidates.
• Benchmarking new forecasting algorithms against established ones on macroeconomic datasets.

Strengths

• The article provides an extension of the model confidence set for multi-horizon comparison of more than two models.
• Simulations demonstrate appropriate size and high power for the proposed tests.
• Empirical illustration is conducted on a large set of macroeconomic variables.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

paperswithcode

Collection Method

Likely contains simulation results and empirical analysis from the associated research article.