A tractable state-space model for symmetric positive-definite matrices

DOI10.1214/14-BA888zbMATH Open1327.62170arXiv1310.5951OpenAlexW1977776321MaRDI QIDQ899053FDOQ899053

Jesse Windle, Carlos M. Carvalho

Publication date: 21 December 2015

Published in: Bayesian Analysis (Search for Journal in Brave)

Abstract: Bayesian analysis of state-space models includes computing the posterior distribution of the system's parameters as well as filtering, smoothing, and predicting the system's latent states. When the latent states wander around $mathbb{R}^n$ there are several well-known modeling components and computational tools that may be profitably combined to achieve these tasks. However, there are scenarios, like tracking an object in a video or tracking a covariance matrix of financial assets returns, when the latent states are restricted to a curve within $mathbb{R}^n$ and these models and tools do not immediately apply. Within this constrained setting, most work has focused on filtering and less attention has been paid to the other aspects of Bayesian state-space inference, which tend to be more challenging. To that end, we present a state-space model whose latent states take values on the manifold of symmetric positive-definite matrices and for which one may easily compute the posterior distribution of the latent states and the system's parameters, in addition to filtered distributions and one-step ahead predictions. Deploying the model within the context of finance, we show how one can use realized covariance matrices as data to predict latent time-varying covariance matrices. This approach out-performs factor stochastic volatility.

Full work available at URL: https://arxiv.org/abs/1310.5951

Recommendations

• Comment on article by Windle and Carvalho
• State space models on special manifolds
• Bayesian semiparametric modeling of realized covariance matrices
• Stochastic covariance models
• A flexible state space model and its applications

zbMATH Keywords

stochastic volatilityrealized covariancebackward sampleforward filter

Mathematics Subject Classification ID

Bayesian inference (62F15) Time series, auto-correlation, regression, etc. in statistics (GARCH) (62M10) Positive matrices and their generalizations; cones of matrices (15B48)

Cites Work

• Generalized autoregressive conditional heteroscedasticity
• Markov chain Monte Carlo methods for stochastic volatility models.
• Title not available (Why is that?)
• DATA AUGMENTATION AND DYNAMIC LINEAR MODELS
• On Gibbs sampling for state space models
• Title not available (Why is that?)
• Bayesian forecasting and dynamic models.
• Selection of Multivariate Stochastic Volatility Models via Bayesian Stochastic Search
• Multivariate Stochastic Variance Models
• Factor Multivariate Stochastic Volatility via Wishart Processes
• The Matrix-Logarithmic Covariance Model
• Econometric Analysis of Realized Volatility and its Use in Estimating Stochastic Volatility Models
• Bayesian analysis of stochastic volatility models with fat-tails and correlated errors
• Title not available (Why is that?)
• The Wishart autoregressive process of multivariate stochastic volatility
• Realized kernels in practise : trades and quotes
• Title not available (Why is that?)
• Multivariate realised kernels: consistent positive semi-definite estimators of the covariation of equity prices with noise and non-synchronous trading
• Local scale models. State space alternative to integraded GARCH processes
• The structure of dynamic correlations in multivariate stochastic volatility models
• Title not available (Why is that?)
• On singular Wishart and singular multivariate beta distributions
• Proof of the conjectures of H. Uhlig on the singular multivariate beta and the Jacobian of a certain matrix transformation
• Bayesian Vector Autoregressions with Stochastic Volatility
• Constructing Stationary Time Series Models Using Auxiliary Variables With Applications
• Multivariate stochastic volatility with Bayesian dynamic linear models
• Forecasting multivariate realized stock market volatility
• Unscented Kalman filtering on Riemannian manifolds
• Bayesian and geometric subspace tracking

Cited In (8)

• Multivariate Stochastic Volatility Model With Realized Volatilities and Pairwise Realized Correlations
• Title not available (Why is that?)
• Bayesian semiparametric modeling of realized covariance matrices
• Comment on article by Windle and Carvalho
• On the Relationship between Uhlig Extended and beta‐Bartlett Processes
• Online learning of time-varying stochastic factor structure by variational sequential Bayesian factor analysis
• Modeling realized covariance measures with heterogeneous liquidity: a generalized matrix-variate Wishart state-space model
• A solution to the positivity problem in the state-space approach to modeling vector-valued time series

This page was built for publication: A tractable state-space model for symmetric positive-definite matrices

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q899053)