The use of extreme value statistics in risk management of the electricity market

In this thesis, we investigate the success of extreme value theory in managing electricity price risk. We specifically deals with the behaviour of the tails of financial time series.The theory provides well established statistical models for which extreme risk measures like the Value at Risk, Expected Shortfall and Return level can be computed. We use daily electricity price data from Nord Pool and compare distributions that effectively estimates the tail quantile. We propose a new method which employs extreme value for estimating the tail risk measure.Our method provides the exact empirical distribution without independence assumption andapplicable to non-stationary data. This method is briefly known as ACER. We show that the recently proposed approach gives better tail quantile estimates, have nice features and it is easy to implement