Imperial College London
Portrait Picture

Emeritus ProfessorNigelMeade

Business School

Emeritus Professor of Quantitative Finance
 
 
 
//

Contact

 

n.meade

 
 
//

Location

 

53 Prince's GateSouth Kensington Campus

//

Summary

 

Publications

Citation

BibTex format

@article{Adcock:2016:10.1016/j.ejor.2016.10.051,
author = {Adcock, CJ and Meade, N},
doi = {10.1016/j.ejor.2016.10.051},
journal = {European Journal of Operational Research},
pages = {746--765},
title = {Using parametric classification trees for model selection with applications to financial risk management},
url = {http://dx.doi.org/10.1016/j.ejor.2016.10.051},
volume = {259},
year = {2016}
}
Download

RIS format (EndNote, RefMan)

TY  - JOUR
AB  - We describe two parametric classification tree methods, which allow formal selection of a member of aclass of generalised distributions. In the paper we consider generalised Beta distributions for non-negativerandom variables and the generalised skew-Student distribution for random variables distributed on thereal line. We introduce a class of symmetric generalised multivariate Student distributions, membersof which may also be selected using the classification trees. We present two versions of the parametricclassification tree: specific to general and general to specific. We apply the classification methods todaily returns on stocks from a selection of 15 major, mid-cap and emerging markets. The results showthat the majority of return distributions follow Student’s t, but that a non-negligible minority follow asymmetric generalised Student distribution. We confirm a well-known stylised fact about skewness: ittends not to be persistent. By contrast, kurtosis is persistent. Using the symmetric generalised multivariateStudent distribution, we present a risk management study based on efficient portfolios constructedfrom UKFTSE250 stocks and specifically concerned with the computation of value at risk. The case studydemonstrates that the model selection procedures based on the classification trees lead to more accuratecomputation of VaR than those based on the normal distribution or on non-parametric approaches. Thestudy also shows that the normal distribution may be used for VaR computations for larger portfolioswhen the holding period is longer.
AU  - Adcock,CJ
AU  - Meade,N
DO  - 10.1016/j.ejor.2016.10.051
EP  - 765
PY  - 2016///
SN  - 0377-2217
SP  - 746
TI  - Using parametric classification trees for model selection with applications to financial risk management
T2  - European Journal of Operational Research
UR  - http://dx.doi.org/10.1016/j.ejor.2016.10.051
UR  - http://gateway.webofknowledge.com/gateway/Gateway.cgi?GWVersion=2&SrcApp=PARTNER_APP&SrcAuth=LinksAMR&KeyUT=WOS:000393932800028&DestLinkType=FullRecord&DestApp=ALL_WOS&UsrCustomerID=1ba7043ffcc86c417c072aa74d649202
UR  - http://hdl.handle.net/10044/1/49201
VL  - 259
ER  -
Download