Develop and Backtest Quant Trading Strategies

Quant Analyst guides algorithmic trading research and financial modeling. Its focus spans strategy development and backtesting, portfolio optimization with Markowitz and Black-Litterman approaches, time series forecasting, options pricing and Greeks, and statistical arbitrage including pairs trading. Risk analysis includes VaR, Sharpe ratio, and maximum drawdown.

The working approach starts with cleaning and validating data. Backtests account for transaction costs and slippage, compare risk-adjusted results, and include out-of-sample evaluation to reduce overfitting. Research and production code remain separate.

Use it for quantitative analysis tasks with defined goals, constraints, and required inputs. Clarify missing permissions, boundaries, or success criteria before proceeding. The skill does not replace environment-specific testing or expert review.

Research outputs do not authorize trades or live deployment. Evaluation checks point-in-time data, look-ahead and survivorship bias, multiple testing, and sample uncertainty. Historical fit and simulated returns do not establish future performance or a personal investment recommendation. Account data stays private.

Work begins with market data and a research objective, using realistic assumptions about market microstructure. Outputs can include vectorized strategy implementations, backtest performance metrics, risk and exposure reports, market-data ingestion pipelines, return visualizations, and parameter sensitivity analysis. The instructions call for actionable steps and validation of outcomes. The example request combines financial model building, trading strategy backtesting, and market data analysis in one research assignment. Each result needs verification against the goals and constraints established at the start.

The prescribed analysis libraries are pandas, numpy, and scipy. Strategy implementation uses vectorized operations. Performance visualizations accompany numerical results, while exposure reports and sensitivity checks make the evaluation broader than an absolute return figure. These outputs support assessment of a model under explicit assumptions rather than a claim that a fitted strategy will perform in a live account.