Columbia University M.A. in Mathematics of Finance
September 2024 – December 2025Coursework: Machine Learning, Time Series Modeling, Stochastic Process, Programming for Computational Finance, Numerical Methods, Derivatives Trading, Option Pricing.
Model. Test. Trade.
End-to-end quant research, from alpha discovery to execution and risk.
Payment
Support my work via PayPal, Venmo or Zelle. Scan or click a QR code to contribute.
Brightness
0%
Font size
14px
Education
University of Liverpool B.Sc. in Mathematics with Finance
September 2020 – May 2024GPA: 3.83/4.0 · First Class Honors (WES report)
Coursework: Stochastic Theory, Linear Statistical Models, Mathematical Risk Theory, Applied Probability, Derivative Securities, Numerical Methods, Operational Research, Statistics and Probability, Financial Reporting.
Completed first two years at Xi’an Jiaotong–Liverpool University, GPA: 3.85/4.0 (Top 5%), then two years at the University of Liverpool as part of the international 2+2 program.
Experience
Numeraxial LLC Quantitative Research Intern
July 2025 – Present- Built a research-grade multi-asset data warehouse with a schema-first design, versioning for cross-asset research.
- Standardized Bloomberg series across IDs, calendars, and FX; built modular ETL for cleaning, corporate actions.
- Engineered documented features and regimes using statistical and macro signals using HMM and clustering.
- Enabled portfolio construction using mean–variance, Black–Litterman, and HRP with constraints.
Western Securities Co., Ltd. Quantitative Analyst Intern
January 2024 – July 2024- Built a research-to-execution stack with a backtester and a pseudo-live setup using the Alpaca Crypto Python API; kept identical signal and state machines; produced deterministic, auditable runs; containerized with Docker, PostgreSQL, and TimescaleDB; added a data-quality guard and a risk worker.
- Implemented two long-only modules: EMA–ADX trend and Z-score reversion gated by ADX and RSI; used conservative GTC limit placement with two-threshold hysteresis; applied affordability and inventory checks.
- Results for a one-year in-sample period: trend-following Sharpe 2.13; reversion Sharpe 2.65; buy-and-hold Sharpe 1.65.
- Built a stress program to define the acceptance envelope before scaling capital: tuned limit offset, reprice thresholds, and freshness gates, and set risk budgets for one-day parametric VaR and drawdown.
- Shipped a Streamlit monitoring dashboard covering P&L, drawdown, turnover, VaR, and an order blotter.
Projects
Predicting Positive SPY Moves with Gradient Boosting CQF
June 2025- Framed next‑day SPY direction as a leakage‑safe classification task; built a reproducible, trading‑calendar–aligned OHLCV pipeline with rigorous timestamp alignment, missing‑data handling, and feature versioning.
- Engineered 40+ economically motivated features including volatility, skew, kurtosis, RSI, MACD, ROC, Bollinger %B, lagged returns, and seasonality; a three‑stage feature funnel—MI → RF‑RFE → GBM.
- Trained XGBoost with time‑series cross‑validation, nested tuning, probability calibration, and stability checks using SHAP and perturbation; achieved a held‑out ROC AUC of 0.714.
- Converted calibrated probabilities into long or flat decisions; evaluated with a walk‑forward procedure.
Portfolio Optimization of Factor‑Based Risk Model using Machine Learning CU
December 2024- Developed a multi‑factor risk model and optimized portfolio weights with factor exposures and covariance matrices.
- Cleaned data with winsorization to mitigate outliers; computed daily residual returns using pseudoinverses.
- Trained regularized least squares and neural networks with cross‑validation to estimate daily factor realizations.
- Backtested on 3,061 stocks with 59 industry factors and 6 style factors (2003–2011); analyzed long/short market values, cumulative P&L, and risk decomposition, achieved consistent risk‑adjusted returns.
Skills & Certificates
Programming
Python (3.x)
C/C++
MATLAB
R
Git
LaTeX
Docker
Jupyter
Excel
Bloomberg
HTML
Languages
- English & Chinese — Bilingual
- Spanish — A2
Certificates
- Certificate in Quantitative Finance (CQF) with Distinction (grades)
- Bloomberg Finance/Market Concepts
Interests
- Chess — Blitz rating 2300
- Piano, Running, Badminton
