基于条件风险值（CVaR）模型的评级公司基金评级质量的研究

发布时间：2018-10-18 19:41

【摘要】：随着基金数量的增多,基金规模的壮大,投资者越来越依赖于第三方评级机构的基金评级。基金评级是信用评级的一个重要支柱,基金评级是客观公正地对基金业绩的评定,并以简单明了的符号形式向社会表达评级结果。对投资者来说,评级提供了基金的有关信息。对发行者来说,综合地反映了基金的绩效和状况。但众多研究和实际结果表明,基金评级不一定能反映基金真实的业绩,且和后续业绩也不一定能保持一致。文章尝试用投资组合常用工具条件风险值(CVaR)来检验基金的评级质量,用风险指标来衡量检验基金。分别建立了基于单目标条件风险值(CVaR)型、多目标条件风险值(CVaR)模型的基金评级质量评价模型。定义了相对准确值、相对总误差以及相对综合指标这三个检验质量的评价指标。然后选取80只开放式基金,综合考察评级公司的基金评级质量。通过数学软件计算出所选基金的条件风险值,按照其大小进行了排名并对其进行了五个星级的评定,最后对比四个评级公司的评级结果进行了实证分析,并用接下来两个月基金的平均收益率对基于条件风险值(CVaR)模型的基金评级结果检验模型的检验能力进行了验证。希望分析结果能为投资者在参考评级公司评级结果时作一个参考。
[Abstract]:As the number of funds increases and the size of the funds grows, investors increasingly rely on the ratings of third-party rating agencies. Fund rating is an important pillar of credit rating. Fund rating is an objective and fair evaluation of fund performance and expresses the results to the society in the form of simple and clear symbols. For investors, ratings provide information about the fund. For issuers, a comprehensive reflection of the Fund's performance and status. But many studies and actual results show that the fund rating does not necessarily reflect the real performance of the fund, and is not necessarily consistent with subsequent performance. This paper attempts to test the rating quality of the fund by using the conditional risk value (CVaR) of the investment portfolio, and to measure the fund by the risk index. Based on the single-objective conditional risk value (CVaR) model and the multi-objective conditional risk value (CVaR) model, the evaluation model of fund rating quality is established. The relative accurate value, relative total error and relative comprehensive index are defined. Then 80 open-end funds are selected to investigate the rating quality of rating companies. The conditional risk value of the selected fund is calculated by mathematical software, and ranked according to its size and evaluated with five stars. Finally, the empirical analysis of the rating results of the four rating companies is carried out. The performance of the model based on conditional risk value (CVaR) model is verified by the average rate of return of the fund in the next two months. It is hoped that the analysis results can be used as a reference for investors when referring to the rating results of rating companies.
【学位授予单位】：浙江工业大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F832.5;F224

【参考文献】