Tests for Principal Eigenvalues and Eigenvectors with S&P 500 Stock Returns

Jianqing Fan's research dataset provides statistical tests for analyzing structural breaks in large factor models. The dataset, last updated in 2026, includes three two-sample tests for evaluating principal eigenvalues, eigenvalue proportions, and eigenvectors. It demonstrates application using daily returns of S&P 500 stocks to analyze events like the 2008 financial crisis and the 2020 pandemic.

Use Cases

• Testing the equality of principal eigenvalues based on divergent spiked covariance models.
• Assessing the stability of eigenvalue proportions to detect shifts in factors' relative importance.
• Evaluating the invariance of principal eigenvectors to identify fundamental factor reorientation.
• Understanding the source of structural breaks in financial factor models based on the described test framework.

Strengths

• Dataset is 378.7 KB, indicating a focused and manageable size for analysis.
• Includes supplementary materials for reproducing the work, as noted in the description.
• Applies methods to a concrete case study of S&P 500 daily stock returns.

Limitations

• Row count and column definitions are unknown, limiting suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset's small size (378.7 KB) suggests limited scope, likely containing test code and results rather than extensive raw data.

Provenance

Source

Jianqing Fan via figshare

Collection Method

Likely contains derived statistical test results and code, based on the description of establishing central limit theorems and developing tests.

Freshness

Last updated 2026-06-04 15:21:53; freshness should be verified.