By employing a multiobjectiveoptimization problem our framework is capable to adapt torisk management practices that are still common in many financial institutionstoday. Thus we avoid the problem of many recent approachesthat try to aggregate different risk metrics that do not fit together. Sincecommon real-world objective functions often incorporate non-linear andnon-convex properties we have chosen a Multi-Objective EvolutionaryAlgorithm (MOEA). The use of the MOEA is illustrated with a realworldportfolio example. Our results support the banks integrated riskmanager in analysing the current position of the bank portfolio withinthe multi-criterial risk/return objective function space. Based on this informationthe manager can derive corresponding mangement decisionsto improve the allocation. Our presented framework is very flexible withrespect to extensions concerning for instance advanced risk measurementmethodologies, correlation assumptions, different time horizons and additionalrisk types.