Worst-case robust decisions for multi-period mean–variance portfolio optimization

摘要: Abstract In this paper, we extend the multi-period mean–variance optimization framework to worst-case design with multiple rival return and risk scenarios. Our approach involves a min–max algorithm for stochastic aspects of scenario tree. Multi-period portfolio entails construction tree representing discretised estimate uncertainties associated probabilities in future stages. The expected value is maximized simultaneously minimization its variance. There are two sources further uncertainty that might require strengthening robustness decision. first some scenarios may be too critical consider terms probabilities. second variance usually inaccurate there different estimates, or either case, best decision has additional property that, return, performance guaranteed view all ex-ante models tested using historical data backtesting results presented.