It isn't a fallacy, it is plausibly optimal if the person being insured is involved in a non-ergodic process. The statistical argument you are applying is too b

It isn't a fallacy, it is plausibly optimal if the person being insured is involved in a non-ergodic process. The statistical argument you are applying is too basic to cover a large number of real-world situations.