Backtesting — Full Backtesting Skill

This skill implements the full 5-stage backtesting methodology from the course material: Data → Research → Metrics → Parameterisation → Validation. It provides:

• 30+ risk/performance ratios (flat, numpy-vectorized, no classes)
• 10 classes of indicators following the course taxonomy (trend-following, oscillators, contrarians, flow, combined, discrete counts, seasonality, statistical, referential, fundamental)
• Event-driven backtesting engine with 8 built-in strategies
• Forward-looking simulation (Johnson SU marginals + t/Gaussian copula)
• Portfolio theory (Markowitz efficient frontier, portfolio-of-portfolios)
• Walk-forward cross-validation with IS/OOS split + gap
• Stress testing with parametric scenario shocks
• Fundamental analysis (Altman Z, Piotroski F, DuPont)

All scripts use only numpy, pandas, and scipy. No heavy dependencies.

Part of the Gauss314 Skills Repository.