Estimator (fynance.estimator)

Maximum-likelihood volatility estimation

fit_volatility fits a GARCH-family conditional-variance model (GARCH / GJR-GARCH / EGARCH(1, 1), with Gaussian or standardized Student-t innovations) on a return series by maximum likelihood. It returns a VolatilityResult carrying the fitted parameters and their standard errors, the information criteria, the in-sample conditional volatility and standardized residuals, plus closed-form / Monte-Carlo variance forecasting (VolatilityResult.forecast) and model simulation (VolatilityResult.simulate).

fit_volatility(y[, model, dist, x0])	Maximum-likelihood fit of a GARCH-family volatility model.
VolatilityResult(params, std_errors, loglik, ...)	Fitted GARCH-family volatility model (see fit_volatility).

The single authoritative conditional-variance recursions and log-likelihood live in fynance.models.econometric_models (loglik_garch); this module only wires the optimiser, the standard errors and the forecasting / simulation on top.

Internal ARMA / GARCH path

Note

The pure-Python ARMA / GARCH parameter estimator (estimation) is an experimental placeholder and is not part of the public API. The public entry point for ARMA / GARCH parameters is get_parameters, documented under Econometric models.