逆周期监管视角下我国开放式基金风险预警模型研究

发布时间：2018-04-15 12:27

本文选题：开放式基金 + 风险预警　；参考：《浙江财经大学》2014年硕士论文

【摘要】：近年来，随着国民经济的快速增长，作为金融业重要组成部分的证券投资基金得到了蓬勃发展，反过来又对国民经济的发展起到一定的积极作用。但是与此同时，基金业内在的风险性也开始逐步暴露出来。作为金融风险监管的一个重要组成部分，对基金业的风险进行合理识别、测度、控制并建立预警模型是有其理论意义与实践意义的。然而，目前理论界对于基金风险管理的研究主要集中于风险的评价与测度；在估算出基金风险之后如何建立有效的预警机制来防范并控制风险还有待进一步的研究。另外，2008年国际金融危机暴露出金融体系的非稳定性以及顺周期性，需要将传统的微观审慎监管理念转向宏观审慎监管理念，从金融系统的顺周期性和逆周期性来研究其对于整个经济体的分散风险或积聚风险的作用。但遗憾的是，大多数学者还是关注金融业的传统行业——银行业，从巴塞尔协议对银行业的资本约束的顺逆周期性来研究金融风险的稳定性，很少有学者涉及证券业，而从基金业顺逆周期性来研究基金风险监管方法的更是少之又少。因此，本文试图从基金业顺逆周期性的角度出发，建立逆周期风险预警模型。本文首先对现阶段已有的国内外研究成果进行一个回顾，总结关于逆周期监管理论和风险预警模型建立的研究成果；其次验证我国开放式基金的顺逆周期性；再次用CHAID决策树和贝叶斯网络来构建我国开放式基金风险预警模型，最后得出结论并给出政策建议。因此，本文的结构安排如下：第一章是绪论，主要介绍本文的研究背景和研究意义，设计研究思路并提出研究创新点；第二章是国内外文献综述，综合讨论目前开放式基金风险研究的不足之处，以及逆周期理论和风险预警模型对其可借鉴之处；第三章介绍开放式基金特征及来源，并将基于中证基金指数的在险值（VaR）与国内生产总值（GDP）、上证综合指数等建立回归模型，检验开放式基金风险的顺逆周期性；第四章建立开放式基金风险预警指标体系，构建基于CHAID决策树算法和贝叶斯网络的风险预警模型并进行实证研究；第五章对全文进行总结与展望，并在逆周期监管理论视角下提出政策建议。实证研究结果表明，我国开放式基金风险与国内生产总值增长率、上证综合指数增长率都存在顺周期性效应，因此建立开放式基金风险的逆周期监管制度非常必要。另外，本文建立了包括16个指标在内的风险预警指标体系，以开放式基金外部机构评级为风险衡量指标，利用CHAID决策树算法和贝叶斯网络构建的开放式基金风险预警模型的检出准确率达到75%~85%，总体预警效果比较不错，基本能够对开放式基金进行准确分类；而且贝叶斯网络模型同时给出各个节点的条件概率，据此可以计算单个个案被分类为某一类的条件概率。本文的创新点有从逆周期监管角度切入来思考开放式基金的风险管理；不仅考虑开放式基金风险测度问题，并验证其顺周期性，为证券业顺周期行为做了一个实证补充，同时考虑了开放式基金的风险预警模型建立问题；在预警模型建立中，，运用数据挖掘中的CHAID决策树算法和贝叶斯网络方法，较以往常用的判别分析法和回归分析法是一个较大的突破。
[Abstract]:In recent years, with the rapid growth of the national economy, the securities investment fund as an important part of the financial industry has been booming, which in turn to the development of the national economy play a positive role. But at the same time, the risk of internal fund industry began to gradually exposed. As an important part of financial risk supervision reasonable, identification, risk of fund industry measure, control and early-warning model has its theoretical significance and practical significance. However, the evaluation and measurement of the present theory for the study of fund risk management mainly focus on the risk; the estimate of fund risk how to establish an effective early-warning mechanism to prevent and control the risks need to be further studied. In addition, the 2008 international financial crisis exposed the stability of the financial system and non cyclical, the traditional micro Prudential supervision concept to the idea of macroprudential supervision, from the pro cyclical and counter cyclical financial system to study the entire economy risk or risk accumulation effect. Unfortunately, most scholars still pay close attention to the financial industry's traditional industries: banking, from Basel Shun protocol to study the stability of financial risk inverse periodic capital constraints on the banking industry, there are few scholars involved in the securities industry, from the fund industry to study the fund risk supervision method of reverse cycle is even less. Therefore, this paper attempts to reverse the cycle of the fund industry's point of view, the establishment of a counter cyclical risk early warning model.
This paper is a review of the existing research results at home and abroad, summarizes the research results based on inverse cycle regulation theory and risk early-warning model; secondly, reverse the periodic verification of China's open-end fund; again with CHAID decision tree and Bayesian network to build China's open-end fund risk early warning model. The final conclusion and gives policy suggestions. Therefore, this paper is organized as follows: the first chapter is the introduction, mainly introduces the research background and significance, research ideas and put forward the design of research and innovation; the second chapter is the literature review at home and abroad, a comprehensive discussion of the current shortcomings of open-end fund risk research, and the inverse cycle theory and the risk early-warning model can be used for reference to it; the third chapter introduces the type and characteristics of open fund sources, and based on the SSE fund index value at risk (VaR) and domestic Gross domestic product (GDP), the Shanghai composite index regression model, reverse the periodic inspection of the open-end fund risk; the fourth chapter establish open-end fund risk early-warning index system, based on the risk early-warning model of CHAID decision tree algorithm and Bayesian network and empirical research; the fifth chapter summary of the article. And put forward the policy recommendations in view of inverse cycle regulatory theory.
The empirical results show that China's open-end fund risk and the GDP growth rate, growth rate of Shanghai Composite Index has pro cyclical effect, it is necessary therefore to establish the inverse cycle regulatory system of open-end fund risk. In addition, this paper establishes the risk early-warning index system including 16 indexes, the open-end fund the external rating agencies for risk measure, using the detection of open-end fund risk early warning model and algorithm of Bayesian networks, decision tree CHAID accuracy rate reached 75%~85%, the overall effect of the early warning is good, can the accurate classification of open-end funds; and the Bayesian network model is given to each node of the conditional probability, which can calculate the individual case is classified as a class of conditional probability.
The innovation of this paper is to consider the risk management of open-end funds from counter cyclical regulatory perspective; not only consider the open-end fund risk measure, and verify its procyclicality, securities industry cyclical behavior adds an empirical, and considers the establishment of risk early-warning model of open-end fund in the early warning; in the modeling, using CHAID decision tree algorithm and Bias network method in data mining is commonly used in the past, discriminant analysis method and regression analysis method is a breakthrough.

【学位授予单位】：浙江财经大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F224;F832.51

【参考文献】