I build models for the losses nobody plans for.
Applied math at UC Irvine, working toward quantitative risk. Below is my work; behind it are notes on everything I study.
GARCH(1,1) returns with Student-t shocks, simulated live. Red is the tail past VaR.
Work
VaR/ES risk engine
A risk engine comparing how different models see the tail of a portfolio's loss distribution, with backtests on each.
Delta-normal, Historical sim, Multivariate-t Monte Carlo, GARCH-FHS, EVT/GPD, Marchenko–Pastur cleaning
StockScreener
A discounted cash flow pipeline that pulls market data and filings, values companies, and ranks them against current prices.
Python, yfinance, DCF valuation
Spearhead
A headless Ubuntu server built from a 2015 MacBook Air that runs network-wide DNS filtering and self-hosted tools.
Ubuntu, Tailscale exit node, Pi-hole v6, cron, PdfDing
Behind the work
Notes on everything I'm studying: risk, probability, optimization, and whatever else I'm working through.
- Sep 20, 2026Filtered historical simulationrisk, time series
- Sep 14, 2026Why expected shortfall replaced VaR in the trading bookrisk
- Sep 2, 2026Cleaning a covariance matrix with Marchenko–Pasturrisk, linear algebra
- Aug 20, 2026Filtered historical simulation in four stepsrisk, time series
Applied and computational mathematics, UC Irvine.