EconStor >
Federal Reserve Bank of New York >
Staff Reports, Federal Reserve Bank of New York >

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

http://hdl.handle.net/10419/60769
  

Full metadata record

DC FieldValueLanguage
dc.contributor.authorGroen, Jan J. J.en_US
dc.contributor.authorKapetanios, Georgeen_US
dc.date.accessioned2010-04-21en_US
dc.date.accessioned2012-08-17T14:32:58Z-
dc.date.available2012-08-17T14:32:58Z-
dc.date.issued2008en_US
dc.identifier.urihttp://hdl.handle.net/10419/60769-
dc.description.abstractWe compare a number of data-rich prediction methods that are widely used in macroeconomic forecasting with a lesser known alternative: partial least squares (PLS) regression. In this method, linear, orthogonal combinations of a large number of predictor variables are constructed such that the covariance between a target variable and these common components is maximized. We show theoretically that when the data have a factor structure, PLS regression can be seen as an alternative way to approximate this unobserved factor structure. In addition, we prove that when a large data set has a weak factor structure, which possibly vanishes in the limit, PLS regression still provides asymptotically the best fit for the target variable of interest. Monte Carlo experiments confirm our theoretical results that PLS regression performs at least as well as principal components regression and rivals Bayesian regression when the data have a factor structure. But when the factor structure in the data is weak, PLS regression outperforms both principal components and Bayesian regressions. Finally, we apply PLS, principal components, and Bayesian regressions to a large panel of monthly U.S. macroeconomic data to forecast key variables across different subperiods. The results indicate that PLS regression usually has the best out-of-sample performance.en_US
dc.language.isoengen_US
dc.publisherFederal Reserve Bank of New York New York, NYen_US
dc.relation.ispartofseriesStaff Report, Federal Reserve Bank of New York 327en_US
dc.subject.jelC22en_US
dc.subject.jelC53en_US
dc.subject.jelE37en_US
dc.subject.jelE47en_US
dc.subject.ddc330en_US
dc.subject.keywordMacroeconomic forecastingen_US
dc.subject.keywordfactor modelsen_US
dc.subject.keywordforecast combinationen_US
dc.subject.keywordprincipal componentsen_US
dc.subject.keywordpartial least squaresen_US
dc.subject.keywordBayesian ridge regressionen_US
dc.subject.stwWirtschaftsprognoseen_US
dc.subject.stwPrognoseverfahrenen_US
dc.subject.stwZeitreihenanalyseen_US
dc.subject.stwFaktorenanalyseen_US
dc.subject.stwBayes-Statistiken_US
dc.subject.stwRegressionen_US
dc.subject.stwUSAen_US
dc.titleRevisiting useful approaches to data-rich macroeconomic forecastingen_US
dc.typeWorking Paperen_US
dc.identifier.ppn587545291en_US
dc.rightshttp://www.econstor.eu/dspace/Nutzungsbedingungenen_US
Appears in Collections:Staff Reports, Federal Reserve Bank of New York

Files in This Item:
File Description SizeFormat
587545291.pdf322.98 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.