High-Dimensional Portfolio Variance and Risk Estimates via Nodewise Regression

Laurent Callot's research article investigates the large sample properties of high-dimensional portfolio variance, weights, and risk. The study uses a nodewise regression technique with the least absolute shrinkage and selection operator to estimate the inverse covariance matrix of excess asset returns. Supplementary materials for the article are available online.

Use Cases

• Estimating global minimum variance portfolio weights based on sparse precision matrix estimation
• Assessing portfolio risk in high-dimensional settings based on inverse covariance matrix estimation
• Comparing portfolio performance between nodewise regression, factor models, and shrinkage methods

Strengths

• The methodology is designed for high-dimensional settings with more assets than observations
• Empirical comparison is performed against factor models and shrinkage methods

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

Research article methodology and empirical analysis