Quantitative statistical robustness of conditional value at risk with moment-based uncertainty set
The quantitative statistical robustness of conditional value at risk(CVaR)in the worst-case is studied.Firstly,CVaR is Lipschitz with random vectors.Secondly,the distribution uncertainty set is locally Lipschitz near a certain point under the Kantorovich metric,thereby the optimal value function of the distributed robust CVaR model satisfies Lipschitz continuity.Finally,the quantitative statistical robustness of the worst-case CVaR is derived.
robust risk measureuncertainty setquantitative statistical robustnessconditional value at risk