Quantative Finance Research

Independent and co-authored quantitative finance research covering machine learning crash prediction, event studies, panel econometrics, and time-series volatility modeling.

---

Featured Projects

Network-Based Early Warning System for Equity Market Crashes

Composite Fragility Score from network biomarkers across 10 S&P 500 sectors, lifting crash-prediction AUC from 0.645 (VIX benchmark) to 0.726 using XGBoost. (Co-authored, under journal review)

Bank Failure Contagion — Event Study (Fama-French Three Factor)

Event study on 37 U.S. bank failures (2015–2025), finding significant negative abnormal returns (t=-3.00***) in geographically proximate banks.

Firm Investment & Volatility Modeling — U.S. Energy Sector

Panel fixed-effects regression on capex drivers (247 firms, R²=94%) plus ARFIMA-GARCH/APARCH volatility modeling on ExxonMobil returns.

Cryptocurrency Market Interdependence — VAR & DCC-GARCH

VAR(1) and DCC-GARCH(1,1) modeling across 5 cryptocurrencies, finding BNB offers the strongest diversification benefit (0.29 correlation vs. 0.74–0.75 among BTC/ETH/DOGE).

GARCH Monte Carlo Risk Simulation (C++)

GARCH(1,1) Monte Carlo simulation implemented from scratch in C++ (1M paths), estimating 1% VaR (7.88%) and Expected Shortfall (9.70%).

---

Additional Projects & Coursework

• Risk Parity Portfolio Construction
• 10-K Textual Analysis & Portfolio Construction
• Betting-Against-Beta Factor Strategy
• Fake News Classification (NLP)