Extreme Value Theory (EVT) has undergone an extensive phase of development in recent years, and it is being applied in an increasing number of areas. Its methods have found particularly fertile ground in quantitative risk management, finance, and insurance. The EVT literature now routinely uses data from these fields to illustrate both theory and practice, and specific issues arising in these contexts have even inspired, and been a focus of, various books such as Embrechts, Klueppelberg, and Mikosch (1997), Beirlant et al. (2004), and Reiss and Thomas (2007). The title of the present book suggests, somewhat misleadingly, a text from which statisticians, financial analysts, or risk managers might learn from scratch how to apply models and methods for rare events. Instead, what we have is a highly technical book that is likely to be inaccessible except to the specialists. Only a small fraction of it is actually devoted to financial applications. The topic is not central to it at all, nor is statistical practice, really. Rather than covering a broad spectrum of extreme value methods, the book is concerned almost exclusively with univariate EVT for dependent sequences of random variables. Within its purview, the book constitutes a valuable addition to the EVT literature. It does a good job of mustering, organizing, and presenting a large and widely scattered corpus of EVT work hitherto unavailable in a textbook form. The bibliography comprises over 400 entries, many of which feature work from the Russian school. The author himself has been a major contributor to this area; his papers account for nearly 10% of the references.
展开▼
