Statistical Modelling of Financial Risk

Preface The business of finance becomes constantly more complex, requiring more advanced statistical tools. Moreover, due to new international regulations, it is more important than ever for financial institutions to understand and measure their risk. The topic of this thesis is to develop new statistical tools for several specific financial applications. The main focus is modelling of risk. The thesis consists of an introduction followed by six papers. Each paper can be read independently of the others. However, the thesis can be divided into two main parts. First a theoretical part, consisting of papers I-III, and then a more application-oriented part consisting of papers IV-VI. Risk management confronts us with heavy-tailed risks, rapid changes and complex interdepen-dencies, which force us to go beyond standard statistical models to develop more sophisticated methodology. Part I treats two themes that are especially important; non-normal distributions (papers I and II) and dependency modelling (paper III). These two statistical themes are very important for a huge set of applications, and in the second part of the thesis, two specific practical problems from finance are treated. In papers IV and V we present a model for measuring total risk for a financial institution, while paper VI develops a joint model for a portfolio of electricity forward products. To summarise, this thesis is a collection of the following six publications: Claudia Czado. Without you, there would be no papers and hence no thesis. Further, it would not have been possible for me to write this thesis if I not had been part of the very stimulating working environment in SAMBA at NR. It is not possible to mention by name everybody in SAMBA I owe credit to, as so many have advised or supported me in one way or another, hence I thank you all. In particular, I wish to thank Line Eikvil and André Teigland, who have been my colleagues and friends at NR for more than 15 years. Line, for teaching me the skills of writing scientific papers and software. My boss André, for giving me lots of freedom and encouragement, and for always being interested in, and enthusiastic about, my work. Further, I am grateful to Håvard Rue, who introduced me to the concept of vines, and also helped Ingrid Hobaek Haff and me writing the skew-t paper. Moreover, he is the one who convinced me to finally put together a PhD thesis after …