ABSTRACT
This chapter introduces a practical customizable framework for the evaluation of portfolio construction approaches. It describes the evolution of techniques from mean-variance optimization and its extensions to strategies that diversify risk across managers. The 2016 study A Simulation-Based Methodology for Evaluating Hedge Fund Investments by Marat Molyboga and Christophe L'Ahelec presented a similar methodology for evaluating portfolio construction approaches that is consistent with investment practices. Portfolio selection has been a fruitful area of research since the introduction of the parsimonious theory, proposed in the 1952 study Portfolio Selection by Harry Markowitz, which reduced a complex asset allocation problem to a simple calculation that relies solely on the vector of expected returns and the covariance matrix. One way to improve the performance of the classic Markowitz mean-variance optimization is to acknowledge that sample means and sample covariance matrices are poor forward-looking estimates and try to improve the quality of estimation by applying Bayesian techniques or shrinkage estimators.
