Mean-Maximum Drawdown Optimization of Buy-and-Hold Portfolios Using a Multi-objective Evolutionary Algorithm

We develop a novel Mean-Max Drawdown portfolio optimization approach using buy-and-hold portfolios. The optimization is performed utilizing a multi-objective evolutionary algorithm on a sample of S&P 100 constituents. Our optimization procedure provides portfolios with better Mean-Max Drawdown trade-offs compared to relevant benchmarks, regardless of the selected subsamples and market conditions. The superior performance of our approach is particularly pronounced in periods with reversing market trends (i.e. a market rally and a fall in the same subsample)