StatGazerBacktests
Reported Sharpe versus reproduced Sharpe
Why a headline performance number is not a review result until timing, universe, costs, and code paths can be regenerated.
Notes
Model review notes for investment teams.
Short technical notes on the questions buyers actually ask before trusting a model, backtest, or quantitative diligence memo.
Start here
Three notes that turn skepticism into a review plan.
The first layer covers the common failure points: the reported backtest that does not reproduce, findings that get blurred into one risk bucket, and model-risk language that is useful even outside banking.
Findings
A finding taxonomy for model validation
How to separate confirmed defects, material assumptions, unresolved questions, limitations, and remediation items.
Governance
SR 11-7-style discipline for non-banks
How to use model-risk vocabulary without pretending a technical review is regulatory assurance.
Reference asset
Use the notes with the public review standard.
The review standard is the checklist behind the notes: decision context, data timing, reproduction, validation design, finding taxonomy, and handoff boundary.