Adjusted Profile Likelihood for Two-Parameter Exponential Family Models

Ferrari, Silvia L. P.; Silva, Michel Ferreira Da; Cribari‐Neto, Francisco

doi:10.1080/03610920509342419

Cited by 2 publications

(1 citation statement)

References 14 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Convém salientar que em pequenas amostras a aproximação da distribuição da estatística de teste LR pela distribuição qui-quadrado pode ser pobre. Nessas situações, podem ocorrer taxas de rejeição da hipótese nula distorcidas (Keblowski, 2005, Ferrari et al, 2005, Stein et al, 2014. Por esse motivo será avaliado o desempenho do teste da razão de verossimilhanças em amostras de tamanho finito, por meio de simulações de Monte Carlo.…”

Section: Teste De Dois Componentesunclassified

Avaliações numéricas das inferências no modelo Beta-Skew-t-EGARCH

Müller

Bayer

2015

Braz. Rev. Fin.

View full text Add to dashboard Cite

The Beta-Skew-t-EGARCH model was recently proposed in literature to model the volatility of financial returns. The inferences over the parameters of the model are based on maximum likelihood method. These estimators have good asymptotic properties, however in finite sample sizes their performance can be poor. With the purpose of evaluating the finite sample performance of point estimators and of the likelihood ratio test proposed to the presence of two components of volatility, we present a Monte Carlo simulation study. Numerical results indicate that the maximum likelihood estimators of some parameters of the model are considerably biased in sample sizes smaller than 3000. The evaluation results of the proposed two-component test, in terms of size and power of the test, showed its practical usefulness in sample sizes greater than 3000. At the end of the work we present an application in a real data of the proposed two-component test and the model Beta-Skew-t-EGARCH.

show abstract

Section: Teste De Dois Componentesunclassified