Optimal sharing, equilibria, and welfare without risk aversion
This paper addresses the impact of empirically observed heterogeneity in individual risk attitudes—particularly loss-domain risk-seeking—on risk-sharing mechanisms, challenging the conventional assumption of universal risk aversion. Method: We develop a general equilibrium model of risk exchange without presupposing risk aversion, employing inverse-monotonic optimization, rank-dependent utility, and expected utility frameworks to characterize Pareto-optimal allocations, existence of competitive equilibria, and validity conditions for the First and Second Welfare Theorems. Contribution/Results: We provide the first rigorous proof of both welfare theorems under pure risk-seeking preferences. Introducing the “jackpot allocation” concept, we identify a scale-dependent mechanism: jackpot allocation is Pareto optimal for small gains but dominated by proportional allocation for large ones—unifying explanations of the disposition effect and small-stakes gambling. Our results resolve a fundamental tension between behavioral evidence and standard general equilibrium theory.