房地产与股票市场： 元回归分析

《房地产与股票市场： 元回归分析》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、REALESTATEANDTHESTOCKMARKET:AMETA-REGRESSIONANALYSISAuthors:DeirdreReilly,BrianLuceyandConstantinGurdgievInstitution:TrinityCollegeDublinABSTRACTTherealestatefinanceliteratureprovidesdiverseandcontradictoryfindingsregardingthedistributionofrealestatere

2、turnsandthelinkagebetweenthesereturnsandstockmarketreturns.Despitetheimportanceofthisrelationshiptotheeconomyingeneralrelativelylittleisknownofwhatcausessuchdifferences.Inthispaper,throughapplyingthetechniqueofmeta-regressionanalysistotheempiricalstudi

3、esintheareaasignificantstepismadetowardsobjectivelyintegratingandsynthesisingtheresultsandidentifyingsystematicvariationsintheresultsofstudies.1.INTRODUCTIONThedistributionalpropertiesofthereturnsofcommonstockshavebeenthesubjectofnumerousempiricalstudi

4、es.AnalysisofkurtosisintheequitymarketdatesbackasfarasMandelbrot(1963)andFama(1965).Forrealestateassets,thereexistsasubstantialbodyofresearchinvestigatingthefirsttwomomentsofreturns(meanandvariance),howeveranalysesofthethirdandfourthmoments(skewnessand

5、kurtosis)isverylimited.Theoverwhelmingmajorityofstudiesthatdoinvestigatethesefourpropertiesfindrealestatereturnstobenon-normallydistributed.Thereislittle,ifany,concreteanalysisofwhatarethefactorsaffectingthedegreeofnon-normality,orinthecasewherereturns

6、arefoundtobenormallydistributed(LizeriandWard,1997;Seileretal.,1999;BrownandMatysiak,2000),whatarethefeaturesofsuchstudies.Despitetheimportanceofrealestateassetstothegeneraleconomy,surprisinglylittleisknownofhowsuchassetsinteractwithothermacroeconomicv

7、ariables.Thereismuchdisagreementinliteratureregardingthenatureoftherelationshipbetweenrealestatepricesandthestockmarket.Thereismixedevidenceastowhethersucharelationshipactuallyexists,andwhereoneisfoundonthesizeanddirectionoftherelationship.Thelackofane

8、xtendedperiodofanalysisandtheomissionofstructuralchangeperiodsinexistingliteraturehasledtoconfusionregardingthenatureofanyrelationship.Thepurposeofthispaperistoovercomemanyoftheselimitationsbyemployingmeta-regressionanalysistointegratea