Optimal Versus Naive Diversification : False Discoveries, Transaction Costs And Machine Learning

This paper shows that sophisticated diversification strategies outperform the 1/Nrule when adjusting for multiple testing; however, their edge is severely underminedby transaction costs. As a way forward, this paper provides a machine learning approachfor ex-ante strategy selection. By linking the characteristics of investment scenariosto the out-of-sample performance of strategies, the algorithm never underperformsthe 1/N rule, even in the presence of relatively high transaction costs