Foundations of Generative AI Trading

This unit establishes core skills for algorithmic trading and asset management with generative AI: acquiring and preprocessing market data, implementing portfolio analytics, translating strategy specifications into backtest code, and introducing machine learning and generative-model concepts. Students will use Python/MATLAB tools to compute risk metrics, construct efficient frontiers, run backtests, and apply LLM-based sentiment analysis to macro speeches, preparing them for advanced model development and deployment in subsequent units.

Learning Objectives

• Analyze historical market data by retrieving adjusted closing prices, cleaning time series, computing daily returns, and calculating asset-level Sharpe ratios using Python or MATLAB.
• Apply programming tools and environments to plot historical risk-free rates, construct mean-variance efficient frontiers, and identify the tangency (optimal) portfolio.
• Demonstrate the ability to translate a written trading strategy specification into executable backtest code, run backtests, and summarize performance metrics and results.
• Evaluate machine learning and generative-model concepts by comparing supervised and unsupervised approaches, describing diffusion and representation-complexity ideas, and applying LLM-based sentiment analysis to a Fed press conference to derive potential trading signals.