Please use this identifier to cite or link to this item: http://hdl.handle.net/123456789/6043
Full metadata record
DC FieldValueLanguage
dc.contributor.authorAmponsah, Irene Kafui Vorsah-
dc.contributor.authorAcquah, Henry De-Graft-
dc.contributor.authorHoward, Nathaniel Kwamena-
dc.date.accessioned2021-09-06T10:13:43Z-
dc.date.available2021-09-06T10:13:43Z-
dc.date.issued2019-
dc.identifier.issn23105496-
dc.identifier.urihttp://hdl.handle.net/123456789/6043-
dc.description6p:, ill.en_US
dc.description.abstractThe Minimum Description Length (MDL), a less known criterion, is making great strides in model selection as compared to the widely known and used information criteria (AIC, BIC, etc). This study developed the MDL criterion using R-functions to evaluate Asymmetric Price Transmission (APT) models (Complex, Standard and Houck’s) for the first time ever. All six criteria’s ability to recover the true DGP was assessed under the condition of varying sample size. A 1000 Monte Carlo simulation procedure revealed that the MDL criteria on the average points to the true DGP and are comparable (if not better) to both AIC and BIC under study condition. Generally, the performances of all model selection criteria (rMDL, nMDL, gMDL, eMDL, AIC and BIC) improved with increasing sample size in their ability to recover the true DGP for both standard and complex models. This study recommends the use of MDL criterion in model selection and in the light of constraint (financial, time and inadequate resources), a sample size of 150 is sufficient in making sound decisions on asymmetric price modelsen_US
dc.language.isoenen_US
dc.publisherUniversity of Cape Coasten_US
dc.subjectModel selectionen_US
dc.subjectMinimum Description Lengthen_US
dc.subjectMonte Carlo Simulationen_US
dc.subjectAsymmetric Price Transmissionen_US
dc.subjectRecovery rateen_US
dc.subjectTrue data generating processen_US
dc.subjectSample sizeen_US
dc.titleComparative Evaluation of Asymmetric Price Transmission Linear Models Using rMDL, eMDL, nMDL, gMDL, AIC and BIC Across Varying Sample Sizesen_US
dc.typeArticleen_US
Appears in Collections:Department of Mathematics & Statistics



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