Extreme Events and Portfolio Construction [26]

Go to: Summary | Previous | Next   
Bullet points include: Output results are notoriously sensitive to input assumptions.Treat quantitative models with scepticism? Focus on reverse optimisation? Techniques proposed to tackle this issue include: Robust approaches and Bayesian priors/anchors, e.g. Black-Litterman. Shrinkage. Resampled optimisation. Essentially all suffer from the ‘fine structure’ problem: the fine structure of optimised portfolio inherently depends on practitioner’s (or model creator’s) subjective views (or, for e.g. Black-Litterman, how these views are expressed)

Contents | Prev | Next | Library

Desktop view | Switch to Mobile