assess-covariance-method

Validate covariance scoring rules by benchmarking them against a registry of established methods using out-of-sample synthetic and real data, ensuring they conform to non-negotiable statistical invariants before reporting results.

Benchmark covariance methods by running them against a registry of established estimators within the precise library, validating performance out-of-sample and checking conformance to non-negotiable invariants for honest reporting.

Perform statistical inference on dependent loss differentials by following the protocol outlined in the SKILL.md file, which provides guidance on pairwise significance testing and accurate reporting of covariance method assessments.

Yes, you can use Python for covariance method conformance testing by installing the precise library with pip install precise[research], which provides the required statistical computation and simulation environment.

Evaluating covariance methods without checking non-negotiable invariants risks inaccurate and dishonest reporting, as estimators may fail fundamental statistical properties needed for reliable benchmarking and out-of-sample validation.