| Texto completo | |
| Autor(es): |
Laurini, Marcio P.
;
Coli de Souza Monteneri Nacinben, Joao Pedro
;
Franco, Joao Pedro M.
;
Chaim, Pedro
Número total de Autores: 4
|
| Tipo de documento: | Artigo Científico |
| Fonte: | JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION; v. N/A, p. 46-pg., 2025-10-10. |
| Resumo | |
We leverage Bayesian inference with Integrated Nested Laplace Approximations (INLA) to develop multifactor models that integrate distinct stochastic processes for the latent factors, resulting in significant improvements in modelling multivariate stochastic volatility. Our approach provides a computationally efficient framework for capturing complex volatility dynamics. The model allows latent factors to exhibit diverse dynamic behaviours, including stationary autoregressive processes, first- and second-order random walks, and long-memory processes. This flexibility enables the model to capture both short- and long-memory dynamics, as well as the impact of shocks with permanent effects on the conditional variance structure. By incorporating these methods, our framework achieves scalability and robustness in estimating model parameters, latent factors, and predictive distributions. Our results underscore the importance of incorporating diverse latent factor dynamics, especially long-memory processes, in improving model fit and out-of-sample forecast. (AU) | |
| Processo FAPESP: | 23/02538-0 - Séries temporais, ondaletas, dados de alta dimensão e aplicações |
| Beneficiário: | Aluísio de Souza Pinheiro |
| Modalidade de apoio: | Auxílio à Pesquisa - Temático |