Estimation of limiting conditional distributions for the heavy tailed long memory stochastic volatility process (Q2375847): Difference between revisions

Consider a stochastic volatility process defined as \[ Y_{i}=\sigma\left( X_{i}\right) Z_{i},\mathstrut i\in\mathbb{Z}\text{,} \] where \(\sigma\) is some (possibly unknown) positive function, \(\left\{ Z_{j},j\in\mathbb{Z}\right\} \) is an i.i.d. sequence and \(\left\{ X_{j} ,j\in\mathbb{Z}\right\} \) is a stationary Gaussian process with mean zero, unit variance, and independent from the process \(\left\{ X_{j}\right\} \). The authors study certain extremal properties\ of the finite dimensional joint distributions of the process \(\left\{ Y_{j}\right\} \) when \(Z_{1}\) is heavy tailed and the Gaussian process \(\left\{ X_{j}\right\} \) possibly has long memory. For fixed positive integers \(h<m\) and \(h^{\prime}\geqq0\), Borel sets \(A\subset\mathbb{R}^{h}\) and \(B\subset\mathbb{R}^{h^{\prime}+1}\), they are interested in the limits \[ \rho\left( A,B,m\right) =\lim\limits_{t\rightarrow\infty}\mathbb{P}\left( \left( Y_{m},\dots,Y_{m+h^{\prime}}\right) \in B\mid\left( Y_{1} ,\dots,Y_{h}\right) \in tA\right) \text{.} \] The general aim of this paper is to investigate the existence of the limiting conditional distributions \(\rho\left( A,B,m\right)\) and their statistical estimation. The asymptotic properties of estimators are studied.