A Mixture-of-Normal Distribution Modeling Approach in Financial Econometrics: A Selected Review

Abstract

This paper provides a selected review of the recent developments and applications of mixture-of-normal (MN) distribution models in financial econometrics. One noted feature of the MN model is its flexibility in accommodating various shapes of continuous distributions, and its ability in capturing leptokurtic, skewed and multimodal characteristics of financial time-series data. The survey is conducted under two broad themes: (i) estimation methods and (ii) applications to financial econometrics.

Keywords: Mixtures of Normal, Maximum Likelihood, Moment Generating Function, Characteristic Function, Switching Regression Model, (G)ARCH Model, Stochastic Volatility Model, Autoregressive Conditional Duration Model, Stochastic Duration Model, Value at Risk

JEL Classification: C22, C53, G19