Yes / This is a pioneering effort to test the comparative performance of two competing models for out-of-sample forecasting the term structure of volatility of crude oil price changes employing both symmetric and asymmetric evaluation criteria. Under symmetric error statistics, our empirical model using the estimated growth factor of volatility through time is overall superior, and it beats in most cases the benchmark model of the square-root-of-time for holding periods between one and 250 days. Under asymmetric error statistics, if over-prediction (under-prediction) of volatility is undesirable, the empirical (benchmark) model is consistently superior. Relative performance of the empirical model is much higher for holding periods up to fifty days.

Submitted by Andre Akamine (andre_akamine@yahoo.com.br) on 2014-02-25T19:51:22Z No. of bitstreams: 1 dissertacao-andre_akamine -versao final.pdf: 1090370 bytes, checksum: 2cc992eb83cbdbf42457a77eeb02dffa (MD5) / Approved for entry into archive by Vera Lúcia Mourão (vera.mourao@fgv.br) on 2014-02-25T20:48:28Z (GMT) No. of bitstreams: 1 dissertacao-andre_akamine -versao final.pdf: 1090370 bytes, checksum: 2cc992eb83cbdbf42457a77eeb02dffa (MD5) / Made available in DSpace on 2014-02-26T12:20:22Z (GMT). No. of bitstreams: 1 dissertacao-andre_akamine -versao final.pdf: 1090370 bytes, checksum: 2cc992eb83cbdbf42457a77eeb02dffa (MD5) Previous issue date: 2014-01-29 / Com o objetivo de analisar o impacto na Estrutura a Termos de Volatilidade (ETV) das taxas de juros utilizando dois diferentes modelos na estimação da Estrutura a Termo das Taxas de Juros (ETTJ) e a suposição em relação a estrutura heterocedástica dos erros (MQO e MQG ponderado pela duration), a técnica procede em estimar a ETV utilizando-se da volatilidade histórica por desvio padrão e pelo modelo auto-regressivo Exponentially Weighted Moving Average (EWMA). Por meio do teste de backtesting proposto por Kupiec para o VaR paramétrico obtido com as volatilidades das ETV´s estimadas, concluí-se que há uma grande diferença na aderência que dependem da combinação dos modelos utilizados para as ETV´s. Além disso, há diferenças estatisticamente significantes entre as ETV´s estimadas em todo os pontos da curva, particularmente maiores no curto prazo (até 1 ano) e nos prazos mais longos (acima de 10 anos). / For the purpose of analyzing the impact in Volatility Term Structure (VTS) of interest rate using two different models in the estimation of the Term Structure of Interest Rates (TSIR) and the assumption regarding the heterocedastic structure of errors (OLS and GLS weighted by duration), the technique proceeds in estimating the VTS using the historical volatility by the standard deviation and autoregressive model Exponentially Weighted Moving Average (EWMA). Through the backtesting test proposed by Kupiec for parametric VaR obtained with the volatilities of VTS’s estimate, conclude that there is a big difference in adherence that depend on the combination of the models used for VTS’s. In addition, there is statistically significant differences between the VTS’s estimated around the points of the curve, specially higher in the short term (less than 1 year) and long term (over 10 years).