Analytical VaR for Nord Pool electricity derivatives : an adjusted riskmetrics approach

Abstract

VaR has emerged as the industry standard for risk reporting, applicable for virtually all corporations which are exposed towards market prices, but is especially relevant for banks and other speculative parties which are regulated under the Basel II accord. The Nord Pool electricity derivative market is extensively applied by hedgers as well as speculators, thus the necessity of estimating VaR for portfolios including such contracts. The model presented in this thesis is based upon the RiskMetrics approach, but is ultimately somewhat adjusted due to the special characteristics of the electricity markets. Moreover, because the calculations and amount of data required for this thesis are extensively, it has resulted in the development of an application written in C#.Net, complemented by SQL commands for easier and faster calculations. A great deal of the workload of this thesis has been in the development of this application. The validity of the model has been examined by back testing 12 real-world portfolios over year 2009 as the sample period. The null hypothesis of the back test is that the expected exception ratio is equal to the actual exception ratio. The results of the back test has been failure to reject the null hypothesis for any of the 12 real-world portfolios, thus this thesis cannot present any statistical evidence that the model is faulty.