Median-adaptive portfolios: a minimum criteria approach to asset allocation

Abstract We propose a new class of adaptive portfolios for asset allocation, based on one-parameter variation the equally weighted portfolio and use median-ranked asset. Our methodological contribution offers simple way performing, static or optimized, allocation assets in any dimension, thus easily circumventing “curse dimensionality”. results show that, even selection parameter that defines our we obtain improved performance compared to benchmark all standard metrics. For case an optimized offer from minimum variance optimization, do require estimation covariance matrix, but approach can be adapted other kinds objective functions. This added to, as complement substitute, existing method.