Risk decision support system for asset liability management

This paper outlines a risk decision support system designed to arrive at well-substantiated policy decisions using asset liability management (ALM) models. The risk decision support system explicitly takes into account that 1) there are multiple risk and returnmeasures that are all important to some extent, 2) there are multiple stakeholders with different (and typically to some extent onflicting) objectives and 3) there is model uncertainty and therefore (economic) stress tests should also play a role in the decision making process.The method is based on techniques from the operations research literature on multicriteria decision making and group decision support to map outcomes generated by any ALM model into policy preferences of (the stakeholders of) the organization in question. Literature on robust control theory is used to guide us how to make good decisions when the ALM model is regarded as just an approximation to an unknown and unspecified model that actually generates the data.The risk decision support system can be used theoretically to obtain insight in for example the influence of a change in future regulation on policy choices or to help design future (pension) contracts. In practice experience learned us that the method facilitates discussions and decision making substantially since the approach helpsparties to focus on interests (objectives), rather than positions (alternatives).