The financial business has become more complex due to new financial instruments, global market information quickly available, and new risk structures, requiring advanced statistical tools. Further, pressure from shareholders and international changing regulatory environment, have made it more important than ever for banks and insurance companies to understand and evaluate their risk exposures.
In banking, the Basel II international capital framework demands statistical methods. The evolution of increasingly complex financial products, such as collateralised debt obligations and credit derivatives, underlines the need for improved knowledge and practice related to risk. Similar demands arise from the new reporting standards IFRS in accountancy, which requires mathematically sound valuation of derivatives, options and intangibles. In insurance, Solvency II introduces a need for new methods to quantify risk over time, beyond traditional actuarial models. Our main focus is risk modelling. The innovation aim in this area is to design new statistical tools for several specific financial applications. The corporate partners consist of a financial group, a non-life insurer and an oil and energy company. These are exposed to market risk (fluctuations in interest rates, currencies and equity) and operational risk (the risk of loss due to ineffective internal systems, or external events such as natural disasters and criminal acts). In addition, credit and insurance risk are crucial to banks and insurers, respectively, while commodity price and production risk are the most important for oil and electricity companies.
Risk management confronts us with heavy-tailed risks, rapid changes and complex dependencies. This forces us to look beyond standard statistical models and simplifying assumptions, to develop more sophisticated methodology. Two themes are especially important, non-normal distributions and dependency modelling. The distributions of financial returns, oil and energy prices, operational losses and insurance claims all have multivariate heavy tails. Moreover, some of them are skewed, having some tails heavier than the others. We will produce new models for heavy-tailed and skewed phenomena. Appropriate dependence modelling is very important. Examples are pricing of credit derivative products, referencing a portfolio of underlying assets, understanding the relationships among different lines of business, and maximizing the profit of a power-generating plant. Dependence structures are often non-linear, requiring new statistical methods, like copula-based approaches. We will develop new ways to construct multivariate distributions from smaller components, building a theory beyond conditional independence. This is particularly useful when dealing with extreme behaviour. These statistical themes are important for many situations, in Norway and worldwide. Hence, any venture that can establish an early lead in this new field will be likely to make an international impact.
The corporate partners are DNB, Gjensidige and Norsk Hydro ASA. DNB is the largest financial institution in Norway and one of the largest in the Nordic region. Gjensidige is Norway's largest non-life insurance company. Norsk Hydro ASA is one of the largest companies in Norway and a world leading aluminium producer.
The day of finance - Thursday October 20, 2011.
The presentations of "Finansdagen" (all slides are in Norwegian):
- Frigessi, Arnoldo: Innledning.
- Ferkingstad, Egil: Grafisk modell for sammenhengen mellom råvarepriser
- Günther, Clara-Cecilie: Modell for å predikere kundeavgang
- Løland, Anders: Justering av korrelasjonsmatriser
- Haug, Ola: Klima og forsikring
- Aas, Kjersti: Copula-modell for markedsrisiko