NEW YORK · COLUMBIA MATHEMATICS OF FINANCE · OPEN TO 2026 ROLES
Data-driven answers to market questions.
I'm Peter Xu — quantitative research at ask2.ai, wealth management at HSBC, equity research at UBS. This site is a working portfolio: every chart on it is produced by Python, SQL, and financial models I wrote, running on real market data.
Selected work
Signal Research & Backtesting
Cross-sectional momentum and mean-reversion signals backtested on 25 US large caps over 10 years — equity curves, drawdowns, parameter sensitivity, and one honest negative result.
PYTHON · PANDAS · NUMPY · BACKTESTING
Interactive DCF: Tesla
A five-year FCFF model on Tesla's real filings with a CAPM-derived discount rate. Drag the assumptions and watch fair value, upside, and the sensitivity grid re-price.
VALUATION · DCF · REACT · EQUITY RESEARCH
SQL Market Analytics
Five analytical questions answered in pure SQL over a relational price database — momentum ranks, sector matrices, rolling volatility, drawdowns, a crossover scanner.
SQL · SQLITE · WINDOW FUNCTIONS
Experience
- Built and tested data-driven trading signals on historical market data, evaluated on return, volatility, drawdown, and risk-adjusted metrics
- Tested signals, factors, and prediction models on US equities and options in Python, R, and SQL; applied NLP to financial news and trade signals
- Supported risk assessments for ultra-high-net-worth clients and designed dynamic asset-allocation tools for structured notes and snowball options
- Developed index-futures arbitrage strategies, optimizing entry/exit thresholds on minute-level spread data
- Analyzed NEV technology pathways and competitive landscapes, quantifying impacts of price cuts, subsidy rollbacks, and dual-credit policy
- Built dynamic DCF models with sensitivity testing for sector coverage
- Helped manage $250,000+ of student-body funds; led research in the consumer cyclical and defensive sectors
- DCF, WACC, comparable-company, and regression modeling