Why the variance? (Q1265985)

The second moment of a random variable attains the minimum value when taken around the mean of the random variable, and this property is characteristic for the second moment among all continuous measures of dispersion. The present paper extends this fact to the multivariate case, derived from the more general following result: If any bounded random vector is a Lehmann unbiased estimator of its mean vector with respect to a continuous loss function \(\varphi\), then \(\varphi\) is necessarily of quadratic form.