Chapter 20. Risk Assessment and Portfolio Construction
JARROD W. WILCOX, PhD, CFA
President, Wilcox Investment Inc.
Abstract: Markowitz mean-variance optimization routines are fundamental to modern risk control and effective diversification. However, their effective use is dependent on accommodating simplifications from the real world. Because of the sensitivity of results to small input variations and of the explosive growth in the number of estimates that must be input as the number of assets is increased, it is useful to take a comprehensive Bayesian view of return and risk estimation. A limitation to risk as variance is not usually a major problem with conventional diversified long-only portfolios but can become so where the impact of skew and kurtosis on the long-run performance of the portfolio is substantial. The nonlinear nature of the optimization process brings benefits to going beyond point estimates and to the use of position constraints. However, too many constraints can seriously hamper the efforts of active investors to get their predictive information through to the realized portfolio.
Keywords: Markowitz mean-variance optimization, portfolio construction, robust portfolio optimization, risk management, risk estimation, value at risk, trading costs, Black-Litterman approach, resampling approach, shrinkage, skewness, kurtosis, optimizer, covariance, correlation, Bayesian
Markowitz mean-variance optimization is a great step forward in our understanding of how to achieve ...