01 / THE PRACTICAL QUESTION
A decision, before a model.
A portfolio manager needs a daily loss estimate before deciding how much exposure to keep overnight.
Set a loss threshold, then check how often realized losses cross it.
02 / DATA & COMPARISON
The idea in plain language.
Value at Risk (VaR) marks a loss threshold at a chosen confidence level. Expected Shortfall (ES) looks beyond it and averages the worst losses. A threshold alone cannot tell you how severe those losses are.
- Backtest
- Compare past forecasts with the outcomes that followed them.
Data. Kenneth French daily industry returns through July 2026, with controlled stress simulations.
Baseline. Historical and normal loss estimates under matched windows.
03 / THE EXPERIMENT
What the saved experiment shows.
In the saved 10,000-observation synthetic check, 95% daily VaR is 1.64% and ES is 2.07%. The losses beyond the threshold are more severe than the threshold itself.
Inspect the supporting resultEvidence record: research-validation.json#numerical_checks/01
Explore the related lab04 / RESULTS & LIMITATIONS
Evidence with its boundaries attached.
The related lab is a cross-project demonstration. Read this investigation’s evidence and limits before transferring its conclusions.
Inspect numerical checks and validation records
Loading validation evidence…
Past returns do not identify every future tail event. One-day risk is not a capital requirement.
05 / REPRODUCE
Reproduce and challenge the result.
Code, configuration, and reproduction
The project contains its implementation, configuration, tests, and walkthrough. Download the lab configuration to record the exact parameters used in an interactive run.
Project code and walkthrough