We consider consistent tests for stochastic dominance efficiency at any order of a given portfolio with respect to all possible portfolios constructed from a set of assets. We propose and justify approaches based on simulation and the block bootstrap to achieve valid inference in a time series setting. The test statistics and the estimators are computed using linear and mixed integer programming methods. The empirical application shows that the Fama and French market portfolio is FSD and SSD efficient, although it is mean-variance inefficient.