Methodology
What each model does, why it is here, what it assumes, and where it breaks. Every page ends with results recomputed from the current data, so the claims can be checked.
DeanOS: portfolio models
The models behind the portfolio explorer: performance, risk, volatility, simulation, regimes, factors and stress tests.
Options Pricing
The models behind the options pricer: a closed-form formula, simulation, trees, and the inverse problem of implied volatility.
Transaction ML
The categorizer behind Transaction ML: synthetic data, an explainable text model, evaluation on unseen merchants, and where it fails.
What this is not
These models describe historical data or option prices under stated assumptions. None of them predicts returns, and none of the output is investment advice. The example portfolios are illustrations built from broad, widely held funds and companies, and option prices on this site are model values, not market quotes. The transaction categorizer is trained and tested on invented data only.