Filtering and Smoothing in State-Space Models with Multiple Regimes

A paper and associated materials presenting improved Bayesian filtering techniques and a novel smoother for regime-switching state-space models. The work assesses performance using a New Keynesian DSGE model and three filters, with simulation results showing speed and accuracy improvements. The author is Nigar Hashimzade, and the materials were last updated in April 2026.

Use Cases

• Implementing a computationally stable smoother for latent variables based on the novel recursion described.
• Comparing the performance of GPB1, GPB2, and IMM filters for regime-switching models as demonstrated in the paper.
• Analyzing monetary policy regime shifts in U.S. data from 1947 to 2023 using the proposed IMM filter-smoother pair.

Strengths

• The proposed smoother is described as computationally stable and easy to implement.
• Simulation results show the IMM filter is about three times faster and at least as accurate as the GPB2 filter.
• The smoother further reduces errors by approximately 25% according to the description.

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

figshare

Collection Method

Likely contains simulation results and applied analysis from a research paper.

Time Range

Applied analysis uses U.S. data from 1947 to 2023.

Freshness

Last updated 2026-04-27 16:33:57; freshness should be verified.

Geography

United States