QM017 Hands-on Lab
Scenario-Based Portfolio Stress Testing
Objective
Revalue a baseline and a candidate portfolio on exactly the same synthetic joint scenario cloud and compare terminal loss, VaR, Expected Shortfall, maximum drawdown, and worst-5-day return.
Run
python python/qm017_scenario_stress.pyCheck
The two summaries are produced from one shared scenario tensor. Any portfolio difference is therefore a same-cloud weight/holding effect, conditional on the generator.
Experiment
Replace the candidate with the baseline weights. Every path-level portfolio outcome should become identical. Then change only one weight and inspect the paired terminal-wealth differences.
The simulation is synthetic and is not a claim about expected portfolio performance.