An agent's expected utility depends only on mean and variance
the verdict
REFUTED
the evidence says no
confidence 0/100
An agent's expected utility generally depends on the full probability distribution of payoffs and higher-order moments (such as skewness and kurtosis), unless specific conditions are met—such as quadratic utility or normally distributed outcomes where mean-variance analysis suffices.
Evidence against · 4
Wikipedia: Expected utility hypothesis
cited by 0
Reference article covering this claim directly
See more details
More against · 3
math.stackexchange.com: Agent's expected utility depends only on mean and variance
cited by 0
Shows that a specific quadratic utility function leads to expected utility depending on mean and variance, implying it is not a general rule.
arxiv.org: (untitled)
cited by 0
Notes mean-variance criterion is equivalent to maximum expected utility only under specific conditions like symmetric elliptical distributions.
link.springer.com: A user’s guide to economic utility functions | Journal of Risk and Uncertainty | Springer Nature Link
cited by 0
States higher-order terms like skewness and kurtosis affect expected utility, showing it does not depend solely on mean and variance.