在金融市场中,如何构建最优投资组合来平衡风险和回报是当今研究者所面临的主要问题之一。为了构建最优投资组合,研究者们通常使用的是VaR或CVaR模型。本研究通过综合运用聚类、核密度估计以及分布鲁棒均值-CVaR模型的方法,从而达到提...在金融市场中,如何构建最优投资组合来平衡风险和回报是当今研究者所面临的主要问题之一。为了构建最优投资组合,研究者们通常使用的是VaR或CVaR模型。本研究通过综合运用聚类、核密度估计以及分布鲁棒均值-CVaR模型的方法,从而达到提升股票投资组合的构建和风险管理能力的目的。本文考虑了包含100只股票日收益数据的实验数据集,通过优化聚类方法,利用核密度估计确定了K-means算法的最佳聚类中心和k值选取。随后,将聚类后的数据输入核密度估计的分布鲁棒均值-CVaR模型中进行分析。通过窗口滚动实验,比较了在有无聚类条件下模型对投资组合收益率的影响。结果显示,应用聚类方法后的模型具有更高的投资组合收益率,有助于投资者更好地平衡风险与回报,构建最优的投资组合。In financial markets, how to construct an optimal investment portfolio that balances risk and return is one of the main challenges faced by researchers today. To build an optimal portfolio, researchers typically use VaR or CVaR models. This study aims to enhance the construction of stock portfolios and risk management capabilities by comprehensively utilizing methods such as clustering, kernel density estimation, and distributionally robust mean-CVaR models. The paper utilized an experimental dataset containing daily returns of 100 stocks. By optimizing clustering methods and determining the optimal clustering centers and k values of the K-means algorithm using kernel density estimation, we then input the clustered data into the robust mean-CVaR model for analysis. By rolling window experiments, we compared the impact of the model on portfolio returns with and without clustering conditions. The results show that the model with clustering methods applied has higher portfolio returns, helping investors better balance risk and return to construct optimal portfolios.展开更多
在供需不确定环境下,企业难以精准地预测供应链上游的供给能力和下游市场的实际需求。在解决如何决策供应商组合和订单分配这一基本问题外,企业还需评估潜在风险并在风险和成本之间寻求平衡点。因此,文中对供需不确定下的供应商选择与...在供需不确定环境下,企业难以精准地预测供应链上游的供给能力和下游市场的实际需求。在解决如何决策供应商组合和订单分配这一基本问题外,企业还需评估潜在风险并在风险和成本之间寻求平衡点。因此,文中对供需不确定下的供应商选择与订单分配问题进行研究,利用均值-条件风险价值(Mean-Conditional Value at Risk,M-CVaR)构建了风险规避的决策模型。数值分析表明:选择合适的供应商数量可以有效降低来自供需两端不确定性的影响;成本随风险规避水平的增加而增加。当风险规避水平较低时,置信水平的变化对决策的影响较小,且增加成本可以显著降低风险。展开更多
园区微能源网在波动的电力现货价格下常面临调度成本的不确定性,易造成额外的损失成本。然而,常用的随机优化手段——典型场景规划、条件风险价值(condition value at risk,CVaR)存在忽略场景合并损失及置信度主观选值的问题。为此,提...园区微能源网在波动的电力现货价格下常面临调度成本的不确定性,易造成额外的损失成本。然而,常用的随机优化手段——典型场景规划、条件风险价值(condition value at risk,CVaR)存在忽略场景合并损失及置信度主观选值的问题。为此,提出兼顾场景相似度与合并损失下的改进场景缩减优化方法,提取典型市场场景集,将场景缩减后的损失度作为置信度的选值依据,形成改进CVaR日前经济调度模型。算例分析表明,基于场景缩减优化引导置信度选值的方法有效促使CVaR值反映实际调度的损失成本,且较主观选值而言更接近理论最低的尾部风险损失,即表明园区微能源网的日前经济调度成本与实际环境更为接近,并进一步讨论了考虑风电不确定性及其他场景缩减方案下的模型推广性。展开更多
Search-based software engineering has mainly dealt with automated test data generation by metaheuristic search techniques. Similarly, we try to generate the test data (i.e., problem instances) which show the worst cas...Search-based software engineering has mainly dealt with automated test data generation by metaheuristic search techniques. Similarly, we try to generate the test data (i.e., problem instances) which show the worst case of algorithms by such a technique. In this paper, in terms of non-functional testing, we re-define the worst case of some algorithms, respectively. By using genetic algorithms (GAs), we illustrate the strategies corresponding to each type of instances. We here adopt three problems for examples;the sorting problem, the 0/1 knapsack problem (0/1KP), and the travelling salesperson problem (TSP). In some algorithms solving these problems, we could find the worst-case instances successfully;the successfulness of the result is based on a statistical approach and comparison to the results by using the random testing. Our tried examples introduce informative guidelines to the use of genetic algorithms in generating the worst-case instance, which is defined in the aspect of algorithm performance.展开更多
文章提出了一种基于长短期记忆网络(Long-short Term Memory network,LSTM)的两阶段均值-CVaR投资组合模型(LSTM+CVaR)。该模型在第一阶段采用LSTM预测股票收益并对股票进行选择;在第二阶段运用均值-CVaR模型来确定所选股票的投资比例...文章提出了一种基于长短期记忆网络(Long-short Term Memory network,LSTM)的两阶段均值-CVaR投资组合模型(LSTM+CVaR)。该模型在第一阶段采用LSTM预测股票收益并对股票进行选择;在第二阶段运用均值-CVaR模型来确定所选股票的投资比例。最后,以沪深300指数股为样本数据,在考虑交易成本和上界约束的情况下,比较LSTM+CVaR模型、LSTM预测选股的等比例模型、随机选股的CVaR模型、随机选股的等比例模型和沪深300指数的风险收益特征、累计收益率和夏普比率。实证结果表明:LSTM+CVaR模型能够实现比传统的投资组合模型更高的平均收益率、收益风险比、累计收益率和夏普比率;减少交易成本和放宽上界约束能提升投资组合模型的表现。展开更多
The article explores a mean-CVaR ratio model with returns distribution uncertainty.To describe the uncertainty of returns distribution,a mixture ellipsoidal distribution absorbing some typical distributions such as th...The article explores a mean-CVaR ratio model with returns distribution uncertainty.To describe the uncertainty of returns distribution,a mixture ellipsoidal distribution absorbing some typical distributions such as the mixture distribution and and ellipsoidal distribution is introduced.Then,by using robust technique with some assumptions,the original robust mean-CVaR ratio model can be formulated as a second-order cone optimization model where the underlying random returns have a mixture ellipsoidal distribution.As an illustration,the corresponding robust optimization models are applied to allocations of assets in securities market.Numerical simulations are presented to illustrate the relation between robustness and optimality and to compare mixture ellipsoidal distribution to some typical distributions as well.展开更多
文摘在金融市场中,如何构建最优投资组合来平衡风险和回报是当今研究者所面临的主要问题之一。为了构建最优投资组合,研究者们通常使用的是VaR或CVaR模型。本研究通过综合运用聚类、核密度估计以及分布鲁棒均值-CVaR模型的方法,从而达到提升股票投资组合的构建和风险管理能力的目的。本文考虑了包含100只股票日收益数据的实验数据集,通过优化聚类方法,利用核密度估计确定了K-means算法的最佳聚类中心和k值选取。随后,将聚类后的数据输入核密度估计的分布鲁棒均值-CVaR模型中进行分析。通过窗口滚动实验,比较了在有无聚类条件下模型对投资组合收益率的影响。结果显示,应用聚类方法后的模型具有更高的投资组合收益率,有助于投资者更好地平衡风险与回报,构建最优的投资组合。In financial markets, how to construct an optimal investment portfolio that balances risk and return is one of the main challenges faced by researchers today. To build an optimal portfolio, researchers typically use VaR or CVaR models. This study aims to enhance the construction of stock portfolios and risk management capabilities by comprehensively utilizing methods such as clustering, kernel density estimation, and distributionally robust mean-CVaR models. The paper utilized an experimental dataset containing daily returns of 100 stocks. By optimizing clustering methods and determining the optimal clustering centers and k values of the K-means algorithm using kernel density estimation, we then input the clustered data into the robust mean-CVaR model for analysis. By rolling window experiments, we compared the impact of the model on portfolio returns with and without clustering conditions. The results show that the model with clustering methods applied has higher portfolio returns, helping investors better balance risk and return to construct optimal portfolios.
文摘在供需不确定环境下,企业难以精准地预测供应链上游的供给能力和下游市场的实际需求。在解决如何决策供应商组合和订单分配这一基本问题外,企业还需评估潜在风险并在风险和成本之间寻求平衡点。因此,文中对供需不确定下的供应商选择与订单分配问题进行研究,利用均值-条件风险价值(Mean-Conditional Value at Risk,M-CVaR)构建了风险规避的决策模型。数值分析表明:选择合适的供应商数量可以有效降低来自供需两端不确定性的影响;成本随风险规避水平的增加而增加。当风险规避水平较低时,置信水平的变化对决策的影响较小,且增加成本可以显著降低风险。
文摘园区微能源网在波动的电力现货价格下常面临调度成本的不确定性,易造成额外的损失成本。然而,常用的随机优化手段——典型场景规划、条件风险价值(condition value at risk,CVaR)存在忽略场景合并损失及置信度主观选值的问题。为此,提出兼顾场景相似度与合并损失下的改进场景缩减优化方法,提取典型市场场景集,将场景缩减后的损失度作为置信度的选值依据,形成改进CVaR日前经济调度模型。算例分析表明,基于场景缩减优化引导置信度选值的方法有效促使CVaR值反映实际调度的损失成本,且较主观选值而言更接近理论最低的尾部风险损失,即表明园区微能源网的日前经济调度成本与实际环境更为接近,并进一步讨论了考虑风电不确定性及其他场景缩减方案下的模型推广性。
文摘Search-based software engineering has mainly dealt with automated test data generation by metaheuristic search techniques. Similarly, we try to generate the test data (i.e., problem instances) which show the worst case of algorithms by such a technique. In this paper, in terms of non-functional testing, we re-define the worst case of some algorithms, respectively. By using genetic algorithms (GAs), we illustrate the strategies corresponding to each type of instances. We here adopt three problems for examples;the sorting problem, the 0/1 knapsack problem (0/1KP), and the travelling salesperson problem (TSP). In some algorithms solving these problems, we could find the worst-case instances successfully;the successfulness of the result is based on a statistical approach and comparison to the results by using the random testing. Our tried examples introduce informative guidelines to the use of genetic algorithms in generating the worst-case instance, which is defined in the aspect of algorithm performance.
文摘文章提出了一种基于长短期记忆网络(Long-short Term Memory network,LSTM)的两阶段均值-CVaR投资组合模型(LSTM+CVaR)。该模型在第一阶段采用LSTM预测股票收益并对股票进行选择;在第二阶段运用均值-CVaR模型来确定所选股票的投资比例。最后,以沪深300指数股为样本数据,在考虑交易成本和上界约束的情况下,比较LSTM+CVaR模型、LSTM预测选股的等比例模型、随机选股的CVaR模型、随机选股的等比例模型和沪深300指数的风险收益特征、累计收益率和夏普比率。实证结果表明:LSTM+CVaR模型能够实现比传统的投资组合模型更高的平均收益率、收益风险比、累计收益率和夏普比率;减少交易成本和放宽上界约束能提升投资组合模型的表现。
基金Supported by the Ministry of Education Planning Fund(Grant No.15YJA790043).
文摘The article explores a mean-CVaR ratio model with returns distribution uncertainty.To describe the uncertainty of returns distribution,a mixture ellipsoidal distribution absorbing some typical distributions such as the mixture distribution and and ellipsoidal distribution is introduced.Then,by using robust technique with some assumptions,the original robust mean-CVaR ratio model can be formulated as a second-order cone optimization model where the underlying random returns have a mixture ellipsoidal distribution.As an illustration,the corresponding robust optimization models are applied to allocations of assets in securities market.Numerical simulations are presented to illustrate the relation between robustness and optimality and to compare mixture ellipsoidal distribution to some typical distributions as well.