以条件风险价值为约束的log-最优投资组合问题

log-optimal Portfolio with a CVaR Constraint

在不允许卖空的情况下,根据Rockafellar提出的条件风险价值,建立了以条件风险价值为风险度量的log-最优投资组合问题,并用Monte Carlo方法和光滑化方法求解这个随机凸优化问题。通过数值试验说明光滑化方法的计算时间比线性化方法的计算时间短,而且分析了投资者的风险厌恶程度对最优投资组合的影响。

Based on the CVaR proposed by Rockafellar,a log-optimal portfolio selection with a CVaR control without the permission of short sale was studied.The stochastic convex optimization was solved via Monte Carlo method and a smoothing method.Numerical experiments showed that the computing time of a smoothing method was less than that of the linear method,and the effect of the investor＇s a risk aversion on optimal portfolio selections was analyzed.