Calibration

stochPathSim.calibration

Calibration of models to time series.

Estimated parameters and the log-likelihood at the optimum.

fit_cir(r, dt)

MLE of dr = kappa (theta - r) dt + sigma sqrt(r) dW with the exact noncentral chi-square transition density, started from OLS estimates.

fit_gbm(prices, dt)

Fit dS = mu S dt + sigma S dW by MLE on log returns.

fit_ou(x, dt)

Exact MLE of dX = kappa (theta - X) dt + sigma dW (also Vasicek).

Uses the AR(1) representation X_{t+dt} = a + b X_t + e.

Maximum-likelihood calibration of diffusion models to time series.

Estimated parameters and the log-likelihood at the optimum.

fit_gbm(prices, dt)

Fit dS = mu S dt + sigma S dW by MLE on log returns.

fit_ou(x, dt)

Exact MLE of dX = kappa (theta - X) dt + sigma dW (also Vasicek).

Uses the AR(1) representation X_{t+dt} = a + b X_t + e.

fit_cir(r, dt)

MLE of dr = kappa (theta - r) dt + sigma sqrt(r) dW with the exact noncentral chi-square transition density, started from OLS estimates.