Abstract: Considering the investors’ psychological characteristics of loss aversion from the perspective of behavioral finance, portfolio optimization models based on linear loss aversion and non-linear loss aversion are constructed. With China stock market data to model static scenario and four dynamic scenarios, this paper empirically studies the optimal asset allocation and performance of different portfolio optimization model with various loss aversion parameters in different scenarios, and then compares results with traditional portfolio selection models. The study demonstrates portfolio optimization models based on loss aversion performance over traditional portfolio selection models, and meanwhile, the asset allocation strategy and portfolio performances are diverse among loss aversion models with different loss aversion parameters and in different scenarios.

Key words: dynamic loss aversion, non-linear loss aversion, portfolio selection, behavioral investment portfolio

摘要： 从行为金融学的角度考虑投资者损失厌恶的心理特征,构建了基于线性损失厌恶和非线性损失厌恶行为投资组合模型。利用中国市场数据模拟一种静态情景和四种动态情景,实证研究不同损失厌恶投资组合模型在不同情景下不同损失厌恶程度的最优资产配置策略和投资绩效表现,并将结果与均值方差模型等传统的投资组合模型进行比较。研究发现损失厌恶投资组合模型优于传统投资组合模型,不同情景下不同程度损失厌恶投资者具有不同的资产配置策略,其投资绩效表现也不尽相同。

关键词: 动态损失厌恶, 非线性损失厌恶, 投资组合, 行为投资组合