Semiparametric tests for behavioral finance efficiency

We propose semiparametric tests for portfolio efficiency, with respect to different Behavioral Finance orderings. In particular, we focus on Markovitz order and Prospect order. We assume that return distributions belong to a scale invariant family, weakly determined by a finite number of parameters: a reward measure, a risk measure and other distributional parameters. We recall stochastic dominance rules for such family of distributions and provide efficiency conditions when the reward measure is isotonic with Markovitz or Prospect type of investors’ preference. Finally, we empirically test portfolio efficiency (in the sense of Markovitz and prospect orderings) when return distribution is uniquely determined by four parameters, using estimation function theory.