Setting default asset allocations for retirement savings accounts can be helpful to individuals who are not experts in finance. However, asset allocations must be made in a way that is appropriate to an individual’s level of risk tolerance and other preference parameters. For example, the issue is not just identifying individuals with relatively high and relatively low risk tolerance; the overall average level of risk tolerance in the population is unclear. The amount of risk a person with an average level of risk tolerance should take is unclear if that average level of risk tolerance is not known.
Recommendations for asset allocation based on solid theory require known the values of risk tolerance, time preference and the elasticity of intertemporal substitution. Existing measures of risk tolerance based on simple hypothetical choices give quantitative measures of risk tolerance, but varying the size of the risks yields measures that are inconsistent. Existing measures of time preference and the elasticity of intertemporal substitution are subject to other inconsistencies. The objective of this proposal is to understand and resolve these inconsistencies by means of structured probing of respondents so that it is possible to give better allocation advice and set better asset allocation defaults.
Miles Kimball will help design surveys and experiments and analyze data resulting from those surveys and experiments. He will also develop theory and help write papers about the reasons why aversion to small risks is so large in relation to aversion to big risks as well as inconsistencies in time preference and intertemporal substitution measures and what these mean for giving advice and setting appropriate defaults.