Predicción en modelos de componentes inobservables condicionalmente heteroscedásticos

Pellegrini, Santiago

Publication:
Predicción en modelos de componentes inobservables condicionalmente heteroscedásticos

dc.contributor.advisor	Ruiz Ortega, Esther
dc.contributor.advisor	Espasa, Antoni
dc.contributor.author	Pellegrini, Santiago
dc.contributor.departamento	UC3M. Departamento de Estadística	es
dc.date.accessioned	2010-03-16T19:52:44Z
dc.date.available	2010-03-16T19:52:44Z
dc.date.issued	2009-04
dc.date.submitted	2009-06
dc.description.abstract	During the last two decades, there has been an increasing interest in the academic and practitioner world on modelling the volatility clustering observed in many economic and financial series. This research was pioneered by Engle (1982) and Bollerslev (1986), with the introduction of GARCH models. It is also common to observe stochastic trends in many economic and financial time series. In this case, a popular practice is to take differences in order to obtain a stationary transformation. Then, an ARMA model is fitted to this transformation to represent the transitory dependence. Alternatively, the dynamic properties of series with stochastic trends may be represented by unobserved component models. It is well known that both models are equivalent when the disturbances are Gaussian. In this case, the reduced form of an unobserved component model is an ARIMA model with restrictions on the parameters; see, for example, Harvey (1989). The main difference between both specifications is that while the ARIMA model includes only one disturbance, the corresponding unobserved component model incorporates several disturbances. Consequently, working with the ARIMA specification is usually simpler. However, using the unobserved components model may lead to discover features of the series that are not apparent in the reduced form model because they arise when estimating the components. When combining both, stochastic trends and volatility clustering, the ARIMA and unobserved component models are not in general Gaussian. This implies that they are no longer equivalent when allowing for conditional heteroscedasticity in the noises. Among the large number of works devoted to studying and applying models that combine these features, almost none of them made a comparative analysis between the two alternatives. This important issue remains somewhat unexplored. Therefore, we think that more effort should be placed in this respect, specially in what regards to forecasting performance. The study of this issue represents the main goal of this thesis
dc.format.mimetype	application/pdf
dc.identifier.uri	https://hdl.handle.net/10016/7385
dc.language.iso	eng
dc.rights	Atribución-NoComercial-SinDerivadas 3.0 España
dc.rights.accessRights	open access
dc.rights.uri	http://creativecommons.org/licenses/by-nc-nd/3.0/es/
dc.subject.eciencia	Estadística
dc.subject.other	Análisis de series temporales
dc.subject.other	Volatilidad
dc.subject.other	Inflación
dc.subject.other	Econometría
dc.subject.other	Modelo econométrico
dc.subject.other	Previsión
dc.subject.other	Finanzas
dc.title	Predicción en modelos de componentes inobservables condicionalmente heteroscedásticos
dc.type	doctoral thesis	*
dc.type.review	PeerReviewed
dspace.entity.type	Publication