EconStor >
Rutgers University >
Department of Economics, Rutgers University >
Working Papers, Department of Economics, Rutgers University >

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10419/59467
  

Full metadata record

DC FieldValueLanguage
dc.contributor.authorKim, Hyun Haken_US
dc.contributor.authorSwanson, Norman R.en_US
dc.date.accessioned2011-06-14en_US
dc.date.accessioned2012-06-25T11:57:18Z-
dc.date.available2012-06-25T11:57:18Z-
dc.date.issued2011en_US
dc.identifier.urihttp://hdl.handle.net/10419/59467-
dc.description.abstractIn this paper, we empirically assess the predictive accuracy of a large group of models based on the use of principle components and other shrinkage methods, including Bayesian model averaging and various bagging, boosting, LASSO and related methods Our results suggest that model averaging does not dominate other well designed prediction model specification methods, and that using a combination of factor and other shrinkage methods often yields superior predictions. For example, when using recursive estimation windows, which dominate other windowing approaches in our experiments, prediction models constructed using pure principal component type models combined with shrinkage methods yield mean square forecast error best models around 70% of the time, when used to predict 11 key macroeconomic indicators at various forecast horizons. Baseline linear models (which win around 5% of the time) and model averaging methods (which win around 25% of the time) fare substantially worse than our sophisticated nonlinear models. Ancillary findings based on our forecasting experiments underscore the advantages of using recursive estimation strategies, and provide new evidence of the usefulness of yield and yield-spread variables in nonlinear prediction specification.en_US
dc.language.isoengen_US
dc.publisherDep. of Economics, Rutgers, the State Univ. of New Jersey New Brunswick, NJen_US
dc.relation.ispartofseriesWorking Papers, Department of Economics, Rutgers, the State University of New Jersey 2011,19en_US
dc.subject.jelG1en_US
dc.subject.ddc330en_US
dc.subject.keywordpredictionen_US
dc.subject.keywordbaggingen_US
dc.subject.keywordboostingen_US
dc.subject.keywordBayesian model averagingen_US
dc.subject.keywordridge regressionen_US
dc.subject.keywordleast angle regressionen_US
dc.subject.keywordelastic net and non-negative garotteen_US
dc.subject.stwPrognoseverfahrenen_US
dc.subject.stwStatistische Methodeen_US
dc.subject.stwWirtschaftsprognoseen_US
dc.subject.stwKonjunkturprognoseen_US
dc.subject.stwTheorieen_US
dc.subject.stwUSAen_US
dc.titleForecasting financial and macroeconomic variables using data reduction methods: New empirical evidenceen_US
dc.typeWorking Paperen_US
dc.identifier.ppn662054202en_US
dc.rightshttp://www.econstor.eu/dspace/Nutzungsbedingungenen_US
Appears in Collections:Working Papers, Department of Economics, Rutgers University

Files in This Item:
File Description SizeFormat
662054202.pdf477.53 kBAdobe PDF
No. of Downloads: Counter Stats
Show simple item record
Download bibliographical data as: BibTeX

Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.