William Shaw, King's College London

Portfolio optimization for general investor risk-return objectives and 
general portfolio distributions

Abstract:
We consider the problem of optimizing a general investor objective (MV, 
Sharpe, VaR, CVaR, Utility, Omega, Behavourial Prospect....) with no 
restrictions on the terminal distributions of the assets comprising a 
portfolio. The solution proposed, initially for long-only portfolios of 
small to modest dimension, is based on introducing an efficient random 
sampling of the simplicial structures characterizing portfolio 
configurations. The sample may be optimized in combination with a 
treatment of risk functions that are either simple analytical objects or 
entities also requiring Monte Carlo simulation of the return 
distribution. Examples will be given. Further details are available
at: ssrn.com/abstract=1680224