Capturing the relevant aspects of phenomena in an econometric model is a fine art. When it comes to the innovation process a trade of between a suitable process and its mathematical implications has to be found. In many phenomena the likelihood of extreme events plays a crucial role. At the same time, classical extreme value theory is based on assumptions that cannot logically be drawn for the phenomenon in question. In this paper, we exemplify the fitness of tempered stable laws to capture both the probability of extreme events, and the relevant boundary conditions in a back-coupled system, the German balancing energy demand.
|Title of host publication||Advances in Directional and Linear Statistics|
|Subtitle of host publication||A Festschrift for Sreenivasa Rao Jammalamadaka|
|Number of pages||16|
|State||Published - 2011|