Risk-value models: Restrictions and applications

Sarin and Weber and others have argued for the expression of preferences under risk by the use of risk-value models, which have the intuitively appealing property that absence of certainty enters the decision-maker's preference function only through a well-defined risk argument. The present paper proposes axioms which should be satisfied by any such model, and explores the restrictions that the axioms place on the preference function. Implications for absolute and relative risk aversion are considered, and the usefulness of the approach is demonstrated with applications to portfolio theory and the theory of the firm.