Submitted by Sara Ribeiro (sara.ribeiro@ucb.br) on 2017-08-17T14:48:23Z
No. of bitstreams: 1
LeonardoNascimentoCastroDissertacao2017.pdf: 1765989 bytes, checksum: fb4cfd563d11e2aa428d7c5af632c835 (MD5) / Approved for entry into archive by Sara Ribeiro (sara.ribeiro@ucb.br) on 2017-08-17T14:50:34Z (GMT) No. of bitstreams: 1
LeonardoNascimentoCastroDissertacao2017.pdf: 1765989 bytes, checksum: fb4cfd563d11e2aa428d7c5af632c835 (MD5) / Made available in DSpace on 2017-08-17T14:50:34Z (GMT). No. of bitstreams: 1
LeonardoNascimentoCastroDissertacao2017.pdf: 1765989 bytes, checksum: fb4cfd563d11e2aa428d7c5af632c835 (MD5)
Previous issue date: 2017-01-01 / Due to the Crisis of 2008, the Basel Committee accelerated the process for update the
Accord and identified some weaknesses such as the inability of V aR to capture the tail risk.
Subsequently, it was recommended to substitute V aR, a non-coherent measure of risk due
to the absence of subadditivity, by CV aR. However, in 2011 the absence of elicitability for
CV aR was shown and this has led some people to believe that it is impossible to perform
a backtesting for this risk measure. Elicitability is an mathematical property for model
selection and not for validation, although the convexity of its scoring function is required
for backtesting. It is important to know the identifiability and testability, which have a
relation with elicitability. For a good backtesting in the Trading Book, the testable function
must be sharp, which is strictly increasing and decreasing with respect to the predictive
and realized variables, respectively, and meet the requirement of ridge backtest, which
depends on the least possible V aR. The CV aR, while not being testable or elicitable, is
at least conditionally elicitable and therefore also conditionally testable. To validate the
CV aR models, simulations were made with the three Acerbi methods, two of this study
for testing and another adapted from the quantile approximation. Of these six, none were
perfect, but two presented better results than the V aR Backtesting. This study analyzed
the risk measures V aR and CV aR by the Historical Simulation, Delta-Normal, Correlated
Normal, Monte Carlo and Quasi-Monte Carlo Simulation methods in the 95%, 97.5% and
99% for the Brazilian bond and stock portfolios, as well as the Brazilian Real against the
Dollar, Euro and Yen currencies, and used some backtesting for the two measures. This
study also proposed a method to improve Backtesting results of V aR. / Devido ?? Crise de 2008 o Comit?? de Basileia acelerou o processo para atualiza????o do Acordo e identificou algumas falhas como, por exemplo, a incapacidade do V aR em captar o risco de cauda. Posteriormente, recomendou-se substituir o V aR, uma medida n??o coerente de risco devido ?? aus??ncia de subaditividade, pelo CV aR. Entretanto, em 2011 foi mostrada a aus??ncia da elicitabilidade para o CV aR e isso induziu algumas pessoas a pensarem ser imposs??vel realizar um backtesting para esta medida de risco. A elicitabilidade ?? uma propriedade matem??tica para a sele????o de modelo e n??o para a valida????o, apesar de que a convexidade de sua fun????o scoring ?? necess??ria para o backtesting. Foram introduzidos os conceitos de identificabilidade e testabilidade, que possuem uma rela????o com a elicitabilidade. Para um bom backtesting no Trading Book, a fun????o test??vel deve ser n??tida, que ?? estritamente crescente e decrescente em rela????o ??s vari??veis preditiva e realizada, respectivamente, e atender o requisito de ridge backtest, que dependa o m??nimo poss??vel do V aR. O CV aR, apesar de n??o ser elicit??vel nem test??vel, ?? pelo menos condicionalmente elicit??vel e, portanto, tamb??m condicionalmente test??vel. Para validar os modelos do CV aR, foram feitas simula????es com os tr??s m??todos de Acerbi, dois desta pesquisa para teste e outro adaptado da Aproxima????o dos N??veis de V aR. Destes seis, nenhum foi perfeito, mas dois apresentaram resultados melhores que o Backtesting do V aR. Esta pesquisa analisou as medidas de risco V aR e CV aR pelos m??todos Simula????o Hist??rica, Delta-Normal, Normal Correlacionado, Simula????o Monte Carlo e Quase-Monte Carlo nos intervalos de confian??a de 95%, 97,5% e 99% para as carteiras de t??tulos e a????es brasileiras, al??m das cota????es do Real frente ??s moedas D??lar, Euro e Iene, e utilizou alguns testes de ader??ncia para as duas medidas. Esta pesquisa tamb??m prop??s um m??todo para melhorar os resultados do Backtesting do V aR.
| Identifer | oai:union.ndltd.org:IBICT/oai:bdtd.ucb.br:tede/2237 |
| Date | 01 January 2017 |
| Creators | Castro, Leonardo Nascimento |
| Contributors | Silva Filho, Osvaldo Candido da |
| Publisher | Universidade Cat??lica de Bras??lia, Programa Strictu Sensu em Economia de Empresas, UCB, Brasil, Escola de Gest??o e Neg??cios |
| Source Sets | IBICT Brazilian ETDs |
| Language | Portuguese |
| Detected Language | English |
| Type | info:eu-repo/semantics/publishedVersion, info:eu-repo/semantics/masterThesis |
| Format | application/pdf |
| Source | reponame:Biblioteca Digital de Teses e Dissertações da UCB, instname:Universidade Católica de Brasília, instacron:UCB |
| Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0017 seconds