Computational Optimization Methods in Statistics, Econometrics and Finance

As outlined in the detailed descriptions of the research teams, their specific expertise and competence make them well suited to contribute to the field of research to be covered by COMISEF. Besides the already existing close ties between some of the partners, the complementarities of the network participants are based on two major aspects:

1. From a technical point of view, it is relevant that the teams contribute specialist knowledge with regard to different aspects of the research topic. While some of the teams are more experienced in the implementation and analysis of optimization tools, in particular heuristic techniques, other teams contribute a profound knowledge in the areas of applications of the techniques, e.g. in applications to macroeconometric modelling or the analysis of financial market data. Only by combining this specific knowledge in the training and research activities of the network, can a major breakthrough be expected resulting in establishing a new optimization paradigm in econometrics and finance based on a careful statistical analysis of methods and their implementations.
2. From an economic point of view, the complementarities between the partners are based on different areas of potential applications covering a wide range from developing new statistical tools over macroeconometric modelling to different applications to financial market data (option pricing, portfolio optimization, etc.). The experience with these different applications is crucial to assess the robustness of the methods in quite differing settings. For a successful introduction as a new optimization paradigm it will be necessary to demonstrate their usefulness beyond some specific applications on a broad bundle of problems including those of real practical interest. With regard to this last aspect, it is important that the network not only includes a partner from industry, but also has well established links of the academic partners with industry.

For the effective training of early-stage researchers in the field of computational methods in statistics, econometrics and finance, appropriate computational equipment is required. In particular, the researchers should have the possibility to implement new methods on parallel and/or distributed computing architecture. Despite of the high cost related to this equipment, the network partners agreed to share existing capacities and to finance additional equipment out of their own resources in order to be able to use the EU funding most efficiently for the research and training activities within the network. In particular, the hardware and software equipment for the researchers recruited for the network will be provided by the local host. Furthermore, the research fellows will be granted access to a parallel computing environment available at the Université de Genève. Finally, NEC Europe Ltd, C&C Research Laboratories (Sankt Augustin, Germany) indicated its readiness to provide computational capacities for research activities of COMISEF.