CMStatistics 2015: Start Registration
View Submission - CMStatistics
B0834
Title: Extremes of Skew-Symmetric distributions Authors:  Boris Beranger - University of New South Wales (Australia) [presenting]
Simone Padoan - Bocconi University (Italy)
Scott Sisson - University of New South Wales (Austria)
Abstract: In environmental, economic or financial fields, the data of real applications can exhibit highly asymmetric distributions. In risk management it is important to analyze the frequency that extreme events such as heat waves, market crashes, etc., occur. Such real processes are high-dimensional by nature. Estimating the dependence of extreme events is crucial for predicting future phenomena, that can have a large impact on real life. A simple way of dealing with asymmetrically distributed data is to use the so-called Skew-Symmetric distributions such as the skew-normal and skew-$t$. These distributions are applied to the context of extreme value theory in order to model the dependence. By defining a non-stationary skew-normal process, which allows the easy handling of positive definite, non-stationary covariance functions, we derive a new family of max-stable processes - the extremal-skew-$t$ process. We provide the spectral representation and the resulting angular densities of the extremal-skew-$t$ process, and illustrate its practical implementation. Finally, an application to wind speed data over 83 locations across the USA is provided.