SVResult.unconditional_variance

property SVResult.unconditional_variance: float

E[exp(h_t)] under the stationary log-normal law.

\(\exp\bigl(\mu + \tfrac12 \sigma^2 / (1 - \phi^2)\bigr)\), the variance of \(y_t - c\) implied by the fit.

Example

>>> import numpy as np
>>> rng = np.random.default_rng(0)
>>> res = SV(rng.standard_normal(200), mean="zero").fit()
>>> expected = np.exp(res.mu + 0.5 * res.sigma2 / (1 - res.phi**2))
>>> bool(np.isclose(res.unconditional_variance, expected))
True