Pilsun Choi

A Comparison of Conditional and Unconditional Approaches in Value-at-Risk Estimation

In this paper, we attempt to find the most important factor causing the differences in the performance of Value-at-Risk (VaR) estimation by comparing the performances of conditional and unconditional approaches. For each approach, we use various methods and models with different degrees of flexibility in their distributions including S-normal distribution, which is one of the most flexible distribution functions. Our empirical results underscore the importance of the flexibility-of-distribution function in VaR estimation models. Even though it seems to be unclear which approach is better between conditional and unconditional approaches, it seems to be clear that the more flexible distribution we use, the better the performance, regardless of which approach we use.

Estimating endogenous switching regression model with a flexible parametric distribution function: application to Korean housing demand

This study introduces Johnsons SU -normal distribution which can accommodate the flexibility of true error distribution to obtain consistent estimates in an endogenous switching regression model. Simulation results indicate that the SU -normal model outperforms the normal model for the consistency of estimators when the error distribution is nonnormal. Korean housing demand model estimated by the SU -normal model also outperforms the normal model in terms of parameter estimates and graphical predictions.

Marginal or copula: which one is critical?

We construct twelve marginal–copula combinations using three marginal distributions (normal, t and SU -normal) and four types of copulas (normal, skewed normal, t and skewed t). Bivariate empirical evidence shows that the choice of marginal distribution plays a more important role in the Value at Risk (VaR) estimation than the selection of copula specification.