File Download
There are no files associated with this item.
Links for fulltext
(May Require Subscription)
- Publisher Website: 10.1016/j.ecosta.2021.10.011
- Scopus: eid_2-s2.0-85119899540
- WOS: WOS:000798794100008
Supplementary
- Citations:
- Appears in Collections:
Article: Multivariate time-series modeling with generative neural networks
Title | Multivariate time-series modeling with generative neural networks |
---|---|
Authors | |
Keywords | ARMA–GARCH model Copulas Dependence Exchange rates Generative moment matching networks Learning distributions Probabilistic forecasts Yield curves |
Issue Date | 2022 |
Citation | Econometrics and Statistics, 2022, v. 23, p. 147-164 How to Cite? |
Abstract | Generative moment matching networks (GMMNs) are introduced as dependence models for the joint innovation distribution of multivariate time series (MTS). Following the popular copula–GARCH approach for modeling dependent MTS data, a framework based on a GMMN–GARCH approach is presented. First, ARMA–GARCH models are utilized to capture the serial dependence within each univariate marginal time series. Second, if the number of marginal time series is large, principal component analysis (PCA) is used as a dimension-reduction step. Last, the remaining cross-sectional dependence is modeled via a GMMN, the main contribution of this work. GMMNs are highly flexible and easy to simulate from, which is a major advantage over the copula–GARCH approach. Applications involving yield curve modeling and the analysis of foreign exchange-rate returns demonstrate the utility of the GMMN–GARCH approach, especially in terms of producing better empirical predictive distributions and making better probabilistic forecasts. |
Persistent Identifier | http://hdl.handle.net/10722/325543 |
ISI Accession Number ID |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Hofert, Marius | - |
dc.contributor.author | Prasad, Avinash | - |
dc.contributor.author | Zhu, Mu | - |
dc.date.accessioned | 2023-02-27T07:34:08Z | - |
dc.date.available | 2023-02-27T07:34:08Z | - |
dc.date.issued | 2022 | - |
dc.identifier.citation | Econometrics and Statistics, 2022, v. 23, p. 147-164 | - |
dc.identifier.uri | http://hdl.handle.net/10722/325543 | - |
dc.description.abstract | Generative moment matching networks (GMMNs) are introduced as dependence models for the joint innovation distribution of multivariate time series (MTS). Following the popular copula–GARCH approach for modeling dependent MTS data, a framework based on a GMMN–GARCH approach is presented. First, ARMA–GARCH models are utilized to capture the serial dependence within each univariate marginal time series. Second, if the number of marginal time series is large, principal component analysis (PCA) is used as a dimension-reduction step. Last, the remaining cross-sectional dependence is modeled via a GMMN, the main contribution of this work. GMMNs are highly flexible and easy to simulate from, which is a major advantage over the copula–GARCH approach. Applications involving yield curve modeling and the analysis of foreign exchange-rate returns demonstrate the utility of the GMMN–GARCH approach, especially in terms of producing better empirical predictive distributions and making better probabilistic forecasts. | - |
dc.language | eng | - |
dc.relation.ispartof | Econometrics and Statistics | - |
dc.subject | ARMA–GARCH model | - |
dc.subject | Copulas | - |
dc.subject | Dependence | - |
dc.subject | Exchange rates | - |
dc.subject | Generative moment matching networks | - |
dc.subject | Learning distributions | - |
dc.subject | Probabilistic forecasts | - |
dc.subject | Yield curves | - |
dc.title | Multivariate time-series modeling with generative neural networks | - |
dc.type | Article | - |
dc.description.nature | link_to_subscribed_fulltext | - |
dc.identifier.doi | 10.1016/j.ecosta.2021.10.011 | - |
dc.identifier.scopus | eid_2-s2.0-85119899540 | - |
dc.identifier.volume | 23 | - |
dc.identifier.spage | 147 | - |
dc.identifier.epage | 164 | - |
dc.identifier.eissn | 2452-3062 | - |
dc.identifier.isi | WOS:000798794100008 | - |