Periodic Ornstein-Uhlenbeck processes driven by Lévy processes (Q4804730)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 1902500
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Periodic Ornstein-Uhlenbeck processes driven by Lévy processes |
scientific article; zbMATH DE number 1902500 |
Statements
Periodic Ornstein-Uhlenbeck processes driven by Lévy processes (English)
0 references
28 September 2003
0 references
periodic processes
0 references
Lévy processes
0 references
Ornstein-Uhlenbeck processes
0 references
0.8387194275856018
0 references
0.7944311499595642
0 references
0.7888922691345215
0 references
0.788605272769928
0 references
A periodic Ornstein-Uhlenbeck process \(X\) with background Lévy process \(Z\) can be defined as the solution of the boundary value problem \[ dX_t=-\lambda X_t dt+dZ_t, t\in[0,1], \qquad X_0=X_1. \] \(X\) is a stationary process (\(P\circ X_t^{-1}=P\circ X_0^{-1}\)), a Markov field (\(\{X_t\), \(t\in[a,b]\}\) and \(\{X_t\), \(t\in[a,b]^c\}\) are conditionally independent given \(X_a\) and \(X_b\)), and, if second moments are finite, its covariance function can be computed explicitly. \(X\) is reversible (\(P\circ X_t^{-1}=P\circ X_{1-t}^{-1}\)) if and only if it is Gaussian. The stationary distribution of \(X\) is the law of \(\int_0^1 e^{\lambda s} dZ_s\). Denote by \(I(\lambda)\) the set of laws of this form. Then, \(\mu\) is the stationary distribution of some periodic Ornstein-Uhlenbeck process with a given parameter \(\lambda\) if and only if \(\mu\in I(\lambda)\). This set depends on \(\lambda\), in contrast with the set of stationary distributions of general Ornstein-Uhlenbeck processes, which is the class of self-decomposable distributions \(\mu\) (for any \(b>1\), there is a distribution \(\mu_{b}\) such that \(\widehat\mu(z)=\widehat\mu(b^{-1}z)\widehat\mu_{b}(z)\), where \(\widehat\mu\) and \(\widehat\mu_b\) are the corresponding characteristic functions), independently of \(\lambda\). In the final section, \(Z\) is taken to be \(Z_t=\kappa W_t\), with \(\kappa>0\) and \(W\) a Wiener process, and the likelihood function of the vector \((\lambda,\kappa)\) is deduced, based on the observation of \(X\).
0 references