α-stable [Alpha-stable] random vectors with time varying spectral measure and applications to financial time series analysis

Autori

• Christoph Hartz

Parametri

• Libri di testo
• Libri di testo universitari

Maggiori informazioni sul libro

This dissertation explores the accurate description of multivariate financial return distributions, which are influenced by time-varying volatilities and dependencies, and often deviate from the Gaussian distribution. A notable alternative is the stable distribution, which accommodates fat tails and skewness. The work introduces various time series models for alpha-stable random vectors, focusing on parameter estimation and practical applications, particularly in portfolio optimization and risk assessment. It generalizes the event study methodology to incorporate heteroscedastic stable error distributions. Additionally, a factor model for stable random vectors with a time-varying spectral measure is developed, allowing parameter estimation through univariate techniques. A simulation method for calculating the density and distribution function of Sub-Gaussian stable random vectors is also presented, facilitating maximum simulated likelihood estimation. Furthermore, two conditional covariation models are introduced to capture the dynamics of the spectral measure of conditional Sub-Gaussian random vectors, serving as natural extensions of Gaussian conditional correlation models. This comprehensive approach enhances the understanding and application of stable distributions in financial contexts.

Lingua: Inglese

Pubblicato: 2008