Concepts of constant absolute risk aversion and constant relative risk aversion have proved useful in the analysis of choice under uncertainty, but are quite restrictive, particularly when they are imposed jointly. A generalization of constant risk aversion, referred to as invariant risk aversion is developed. Invariant risk aversion is closely related to the possibility of representing prefere...

We examine evolutionary basis for risk aversion with respect to aggregate risk. study populations in which agents face choices between and idiosyncratic show that the maximize long-run growth rate are induced by a heterogeneous population least most averse indifferent obtaining its linear harmonic mean sure, respectively. Moreover, an approximately optimal behavior can be simple distribution ...

We consider a portfolio-choice problem with one risky and one safe asset, where the utility function exhibits decreasing absolute risk aversion (DARA). We show that the indirect utility function of the portfolio-choice problem need not exhibit DARA. However, if the (optimal) marginal propensity to invest is positive for both assets, which is true when the utility function exhibits nondecreasing...

W do bidders tend to bid higher than the risk-neutral Nash equilibrium in sealed-bid first price auction experiments? The effect of risk aversion has long been offered as a possible explanation. More recently, several studies proposed regret as another explanation, citing strong experimental evidence. But which effect is more important? We design an experiment to separate the effects of risk av...

This paper studies the relationship between the systematic risk of financial instruments and the volatility of their equilibrium prices in a two-period stochastic asset valuation model. Whereas there is no link between the relative risk of assets and their price volatility in standard representative-agent models with additively-separable preferences, in this model with nonseparable preferences ...

Many mathematical forms of utility functions have been considered in the economics and finance literature, although, as Xie (2000) has said, by far the most widely employed are those displaying constant relative risk aversion (CRRA). Xie pointed out the dangers implicit in an inappropriate assumption of CRRA and the desirability of more flexible utility functions permitting a greater range of r...

We consider an exchange economy under uncertainty, in which agents’ utility functions exhibit constant absolute risk aversion, but they may be recursive and the expected utility calculation may be based on multiple subjective beliefs. The risk aversion coefficients, subjective beliefs, subjective time discount factors, initial endowments, and tradeable assets may differ across agents. We prove ...

The testable implication of the complete risk-sharing hypothesis depends on what is assumed on households' relative risk aversion (RRA) coefficient. We therefore use a hyperbolic absolute risk aversion (HARA) utility function, which includes increasing, constant, and decreasing RRA as special cases, to test this hypothesis. Using household level total non-durable consumption data from Indian vi...