Personal Project
Crisis-Aware Portfolio Risk Model: COVID-19 & 2008 Stress Testing
Built for investors who want to see live, explainable stress tests. The live demo below lets you shock a portfolio with COVID-grade volatility and watch VaR respond.
Live Interactive Demo
Everything runs client-side so you can tweak scenarios without leaving the page. Monte Carlo stress testing updates instantly as you drag sliders.
1. COVID Monte Carlo Stress Test
Stress test portfolio VaR in real time
VaR 95%
0.0%
VaR 99%
0.0%
Expected Shortfall
0.0%
COVID crash line
-34%
The distribution reflects your volatility and correlation settings. The COVID-19 line anchors March 2020’s drawdown; raise correlation to see how tight coupling pulls tails left.
Project Snapshot
- Stress-tested multi-asset portfolios against COVID-19 and 2008-style shocks.
- Blended factor risk with historical drawdown patterns to assess resilience.
- Quantified tail risk via scenario analysis and historical VaR comparisons.
COVID-19’s early-2020 drawdown was a true outlier: even after inflating vol and correlations, the simulated tail still sat above the realized crash—evidence of a one-off shock that normal VaR couldn’t capture.
This second view overlays the 2008 crisis line: both crashes push beyond conventional VaR, with COVID still stretching the tail—showing why stress-tested covariance (high vol, high correlation) is critical for realistic loss sizing.
Methodology
Data & Scenarios
- Historical prices across equities, credit, and alternatives.
- Constructed COVID-19 and 2008 analog shocks to replay tail events.
- Factor loadings to see which exposures amplified risk.
Metrics & Outputs
- Drawdown depth and recovery time comparisons.
- Historical VaR vs. stressed VaR deltas.
- Resilience scorecard summarizing portfolio weak points.
Tech & Tools
Next.js, client-side Monte Carlo simulation, deterministic hydration guardrails, Python backtests.