We believe thatthe following three questions require further investigations.1. In the financial portfolio management context, if decisions need to be takenat multiple stages, is it possible to devise a time consistent risk minimizationpolicy?2. If the value of an asset of interest is strongly in°uenced by a climatic variable(such as temperature or rainfall) that is best modelled by techniques that arenot normally used in financial modelling, how should financial derivativeson such an asset be priced?3. If the undesirable risky phenomenon depends on both the upper and lowertails of a, possibly asymmetric, probability distribution (e.g., that of amountof rainfall), what is a suitable measure of risk in this context?This thesis supplies, at least partial, answers to each of these challengingquestions.