Develop and Backtest Quant Trading Strategies

Provides quantitative analysis expertise for algorithmic trading and financial modeling, following a data-quality-first, risk-adjusted, overfitting-resistant methodology.

Use this skill when developing and backtesting a trading strategy, computing risk metrics for a portfolio, optimizing portfolio allocation, performing time series forecasting on financial data, pricing options and computing Greeks, or implementing statistical arbitrage/pairs trading. Not a fit for non-financial quantitative analysis or for production trading system infrastructure beyond the research/strategy layer.

Focus areas span trading strategy development and backtesting, risk metrics (Value at Risk, Sharpe ratio, maximum drawdown), portfolio optimization (Markowitz mean-variance optimization, Black-Litterman), time series analysis and forecasting, options pricing and Greeks calculation, and statistical arbitrage/pairs trading.

The approach follows five principles: data quality first (clean and validate all inputs before analysis), robust backtesting that accounts for transaction costs and slippage rather than idealized fills, prioritizing risk-adjusted returns over absolute returns, out-of-sample testing to detect and avoid overfitting, and clear separation between research code and production code.

Expected outputs include a vectorized strategy implementation (using pandas/numpy for performance rather than row-by-row loops), backtest results with performance metrics, risk analysis and exposure reports, a data pipeline for market data ingestion, visualizations of returns and key metrics, and parameter sensitivity analysis. Implementations rely on pandas, numpy, and scipy, and incorporate realistic assumptions about market microstructure (bid-ask spreads, liquidity constraints, execution costs) rather than frictionless idealized models.

Built on the Python quantitative stack - pandas, numpy, and scipy - for data manipulation, numerical computation, and statistical analysis. Options pricing and Greeks calculation typically leans on scipy's statistical functions (for Black-Scholes-style closed-form pricing) and numpy for vectorized sensitivity calculations across a range of strikes, expirations, or volatility scenarios rather than looping over individual contracts one at a time.

Quantitative analysts and algorithmic trading developers who need rigorous, overfitting-resistant strategy development and risk analysis rather than naive backtests that ignore transaction costs and market microstructure - particularly useful when a strategy that looks profitable in a simplified backtest needs to be stress-tested against realistic execution assumptions before being trusted with capital.

sharpe_ratio = (returns.mean() - risk_free_rate) / returns.std()