Methodology
Every number in SPZCO is computed by a tested quantlib function and documented here: what it measures, the convention chosen, a textbook reference, and known limitations — in plain language. Methodology transparency is a feature — a wrong number is a Sev-1 bug. These pages explain what each model does and why it can be trusted; precise formulas and calibration parameters (especially for the proprietary engines — fair value, base rates) are not enumerated, and that withholding is disclosed on each page rather than hidden.
The flagship engines
Intrinsic Value (normalized-FCFF DCF)The fair-value engine — a driver-built, normalized-FCFF DCF (with DDM and residual-income routing where the business shape demands it).
Base-Rate Expectations (Mauboussin)Reverse-DCF the price into a required growth rate, then ask history how often companies that size ever sustained it (Mauboussin base rates).
Valuation & signals
Ownership, events & congress
Portfolio & risk analytics
- Active-Risk (Tracking-Error) Decomposition
- Return Attribution (Contribution-to-Return)
- Brinson Sector Attribution
- Multi-period (linked) Brinson attribution
- Up/Down Capture
- Component VaR
- Pairwise Correlation Tools
- Covariance & Correlation
- Distribution Shape & Tail Metrics
- Diversification Ratio & Effective Number of Bets
- Drawdown & Ulcer Index
- Drawdown Metrics & Ratios
- Drawdown Episode Table
- EWMA (RiskMetrics) Covariance
- Additional Performance Ratios
- Multi-Factor Regression
- Risk-Based Portfolio Construction
- Risk-Adjusted Ratios
- Market Model (Beta / Alpha / R²)
- Risk Contribution (MCTR / CTR / PCTR)
- Rolling Trailing-Window Series
- Factor-Shock Scenario / Stress
- Ledoit-Wolf Covariance Shrinkage
- Weights-Mode Return Synthesis
- Value at Risk & Conditional VaR
- Tracking Error & Active Return
- Volatility & Downside Deviation