Can Variance Gamma Distribution parameters be used as Variance Gamma
Process parameters?
The closed form estimations the of Variance Gamma Distribution standard
deviation and kurtosis parameters are
$$\hat\sigma^2=\sum\limits_i X_i^2/N$$
and
$$\hat\nu=\{\sum\limits_i X^4_i/N\}(3\hat\sigma^4)^{-1}-1$$
Can those be used as the $\sigma$ and $\nu$ parameters for the Variance
Gamma Process? If not, do closed form estimations of those parameters from
data exist? If so, what are they?
No comments:
Post a Comment