Working Paper: CEPR ID: DP5177
Authors: Javier Menca; Enrique Sentana
Abstract: We analyse the Generalised Hyperbolic distribution adequacy to model kurtosis and asymmetries in multivariate conditionally heteroskedastic dynamic regression models. We standardise this distribution, obtain analytical expressions for the log-likelihood score, and explain how to evaluate the information matrix. We also derive tests for the null hypotheses of multivariate normal and Student t innovations, and decompose them into skewness and kurtosis components, from which we obtain more powerful one-sided versions. Finally, we present an empirical application to five NASDAQ sectorial stock returns that indicates that their conditional distribution is asymmetric and leptokurtic, which can be successfully exploited for risk management purposes.
Keywords: inequality constraints; kurtosis; multivariate normality test; skewness; student t; supremum test; tail dependence
JEL Codes: C32; C52; G11
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Generalized hyperbolic (GH) distribution (C46) | Accurate representation of financial returns (G17) |
| Generalized hyperbolic (GH) distribution (C46) | Better fit than symmetric student t and normal distributions (C46) |
| Generalized hyperbolic (GH) distribution (C46) | More reliable risk assessments (D81) |
| Distributional assumptions (D39) | Effectiveness of risk management strategies (H12) |
| GH distribution (O40) | Captures higher tail dependence and asymmetries (C46) |
| Traditional models (C59) | Underestimate joint occurrence of extreme events (C46) |