On a threshold heteroscedastic model

Chen, Cathy W. S.; So, Mike K. P.

doi:10.1016/j.ijforecast.2005.08.001

Cited by 186 publications

(85 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…see Chen and Lee (1995) ;Vrontos, Dellaportas, and Politis (2000); Chen and So (2006) and others. MCMC methods simulate iteratively from the conditional posteriors of groups of model parameters.…”

Section: Mcmc Methodsmentioning

confidence: 93%

“…Further details for the other models may be found in Chen and So (2006). The joint posterior distribution for each model is formed by multiplying the likelihood by the joint prior for that model.…”

Section: Gjr-garch Model and Priormentioning

confidence: 99%

“…As such MCMC methods were generally used to forecast VaR thresholds for each model in this paper. We follow Chen and So (2006) and design an efficient, adaptive MCMC sampling scheme for estimation and quantile forecasting.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Bayesian Forecasting for Financial Risk Management, Pre and Post the Global Financial Crisis

et al. 2011

View full text Add to dashboard Cite

Value-at-Risk (VaR) forecasting via a computational Bayesian framework is considered. A range of parametric models are compared, including standard, threshold nonlinear and Markov switching GARCH specifications, plus standard and nonlinear stochastic volatility models, most considering four error probability distributions: Gaussian, Student-t, skewed-t and generalized error distribution. Adaptive Markov chain Monte Carlo methods are employed in estimation and forecasting. A portfolio of four Asia-Pacific stock markets is considered. Two forecasting periods are evaluated in light of the recent global financial crisis. Results reveal that: (i) GARCH models out-performed stochastic volatility models in almost all cases; (ii) asymmetric volatility models were clearly favoured pre-crisis; while at the 1% level during and post-crisis, for a 1 day horizon, models with skewed-t errors ranked best, while IGARCH models were favoured at the 5% level; (iii) all models forecasted VaR less accurately and anti-conservatively post-crisis. ABSTRACT Value-at-Risk (VaR) forecasting via a computational Bayesian framework is considered. A range of parametric models are compared, including standard, threshold nonlinear and Markov switching GARCH specifications, plus standard and nonlinear stochastic volatility models, most considering four error probability distributions: Gaussian, Student-t, skewed-t and generalized error distribution. Adaptive Markov chain Monte Carlo methods are employed in estimation and forecasting. A portfolio of four Asia-Pacific stock markets is considered. Two forecasting periods are evaluated in light of the recent global financial crisis. Results reveal that: (i) GARCH models out-performed stochastic volatility models in almost all cases; (ii) asymmetric volatility models were clearly favoured pre-crisis; while at the 1% level during and post-crisis, for a 1 day horizon, models with skewed-t errors ranked best, while IGARCH models were favoured at the 5% level; (iii) all models forecasted VaR less accurately and anti-conservatively post-crisis.

show abstract

Section: Mcmc Methodsmentioning

confidence: 93%

Section: Gjr-garch Model and Priormentioning

confidence: 99%

See 1 more Smart Citation

Bayesian Forecasting for Financial Risk Management, Pre and Post the Global Financial Crisis

et al. 2011

View full text Add to dashboard Cite

show abstract

“…The Random Walk Metropolis algorithm is used for the first M iterations, the so-called burn-in period, while the Independent Kernel MH algorithm is employed from iteration M + 1 onwards, employing the sample mean and covariance matrix of the burn-in iterates for each parameter grouping. This procedure is discussed in detail in Chen and So (2006).…”

Section: Bayesian Estimation and Forecastingmentioning

confidence: 99%

Forecasting Value-at-Risk using nonlinear regression quantiles and the intra-day range

Chen

Gerlach

Hwang

et al. 2012

International Journal of Forecasting

Self Cite

View full text Add to dashboard Cite

Value-at-Risk (VaR) is commonly used for financial risk measurement. It has recently become even more important, especially during the 2008-09 global financial crisis. We propose some novel nonlinear threshold conditional autoregressive VaR (CAViaR) models that incorporate intra-day price ranges. Model estimation and inference are performed using the Bayesian approach via the link with the Skewed-Laplace distribution. We examine how a range of risk models perform during the 2008-09 financial crisis, and evaluate how the crisis affects the performance of risk models via forecasting VaR. Empirical analysis is conducted on five Asia-Pacific Economic Cooperation stock market indices as well as two exchange rate series. We examine violation rates, back-testing criteria, market risk charges and quantile loss function values to measure and assess the forecasting performance of a variety of risk models. The proposed threshold CAViaR model, incorporating range information, is shown to forecast VaR more efficiently than other models, across the series considered, which should be useful for financial practitioners.

show abstract

“…This is because the correlation between the coefficients is high, and mixing is improved by sampling the volatility coefficients in a block (see for example Chen and So (2006) convergence of these algorithms using two conditions: diminishing adaptation and uniform ergodicity. We do not discuss these concepts further since the methods used in this paper are well studied and convergence is proved in the accompanying references.…”

Section: Updating the Volatility Coefficientsmentioning

confidence: 99%

Modelling the Conditional Distribution of Daily Stock Index Returns: An Alternative Bayesian Semiparametric Model

2011

View full text Add to dashboard Cite

Please cite this publication as follows:Kalli, M., Damien, P. and Walker, S. (2013) Abstract This paper introduces a new family of Bayesian semi-parametric models for the conditional distribution of daily stock index returns. The proposed models capture key stylized facts of such returns, namely heavy tails, asymmetry, volatility clustering, and the 'leverage effect'. A Bayesian nonparametric prior is used to generate random density functions that are unimodal and asymmetric.

show abstract

On a threshold heteroscedastic model

Cited by 186 publications

References 17 publications

Bayesian Forecasting for Financial Risk Management, Pre and Post the Global Financial Crisis

Bayesian Forecasting for Financial Risk Management, Pre and Post the Global Financial Crisis

Forecasting Value-at-Risk using nonlinear regression quantiles and the intra-day range

Modelling the Conditional Distribution of Daily Stock Index Returns: An Alternative Bayesian Semiparametric Model

Contact Info

Product

Resources

About