Asset Allocation under Multi-risk Contrlo Targets

Compared with VaR risk measure,CVaR could measure the portfolio risk more suitable under the sub-additivity condition.And besides,generalized extreme value distribution and copula function could fit the portfolio return rate's fat tails and dependence quite well.In this paper,CVaR and Copula-GEV distribution are adopted to describe the biggest risk,which the portfolio would face with during a sharp drop in the market.We introduce CVaR in Mean-Variance model as a target of risk control.An optimization model of asset allocation under multiple risk targets is proposed,considering the return rate,the volatility and the extreme value risk comprehensively.In accordance with the features of this model,we combine particle swarm optimization method with Monte Carlo simulation to solve it.Through an empirical analysis of stock returns of some listed companies in China,the feasibility and effectiveness of proposed model and algorithm have been validated.