基于VaR多种方法下中国证券投资基金风险度量的研究

发布时间：2018-03-14 00:45

本文选题：证券投资基金　切入点：VaR　出处：《上海师范大学》2013年硕士论文　论文类型：学位论文

【摘要】：我国基金尤其是开放式证券投资基金近年来发展迅速，投资群体、投资规模不断扩大，市场影响力逐步加深。这些迅速壮大的基金尤其是开放式基金不得不面临的问题是如何当我国基金市场对外开放后在全球范围内有效地实现基金资产的保值增值，如何进行基金的风险防范和管理。在我国基金诞生之时国外全新的风险管理工具在险价值（VaR）不断发展壮大起来，现已经成为金融资产的风险管理的主流工具，可以说它的确立是风险管理领域的一次革命，开创了全面风险管理的新时代。文章利用VaR度量当前我国开放式证券投资基金风险值水平，具体使用指数移动平均法、历史模拟法和蒙特卡罗法分别以20天、250天和20天作为时间窗口度量未来一天每支样本基金的潜在损失值，并以2011年1月至2012年9月共21个月作为检验考察每个模型的有效性；另外以GARCH模型计算某一只基金组合的VaR并模型的有效性。最终发现：1、样本基金存在严重的尖峰厚尾性；大部分样本基金不存在ARCH效应。2、基金光大保德信量化核心（360001）等多只基金的风险价值在整个样本基金中相对较大；不能从投资类型来判断基金风险值大小。3、置信水平的选择对模型的有效性有一定的影响。4、使用GARCH-VaR计算出组合在置信度95%和99%的风险价值分别为3.111%和4.400%，该模型在95%置信度下检验结果不显著；5、从模型在样本基金的优劣性看蒙特卡罗模拟法较其他两种方法更优。基于上述结论，投资者或基金公司在使用分析法预测基金的VaR值时要假设与经验分布趋势较一致的分布，对GARCH模型也不能盲目使用；基金公司要考虑选择基金的风险值计算最优模型，，增强风险预测的可靠性，加强对旗下各基金组合风险度量与监测，做好风险防范与管理。
[Abstract]:China's fund especially the open-end securities investment fund has developed rapidly in recent years, the investment group, expanding the scale of investment, market influence gradually deepened. The rapid growth of the fund especially the open-end funds have to face the problem of how to when the fund market in China after the opening in the global scope effectively and increase the value of fund assets and how to conduct risk prevention and management of funds in our country. The birth of the new foreign fund risk management tool value at risk (VaR) of growing up, now has become a main tool of risk management of financial assets, it can be said that the establishment of a revolution in the field of risk management, ushered in a new era comprehensive risk management.
By using the VaR measure the current level of China's open-end securities investment funds risk value, the specific use of exponential moving average method, historical simulation method and Monte Carlo method respectively in 20 days, 250 days and 20 days as a time window measure next day each sample fund the potential loss of value, and from January 2011 to September 2012 a total of 21 months as the effectiveness of the inspection of each model; in addition to effectiveness of GARCH model and VaR model of a fund portfolio. Finally found that: 1, the sample fund has peak thick tail serious; most of the sample funds does not exist in the ARCH effect.2, Baode Everbright fund letter quantify core (360001) value of risk a number of funds in the fund in the whole sample is relatively large; not from the type of investment fund risk value to determine the size of.3, the validity of the confidence level selection for model has certain effect on.4, GARC H-VaR calculates the portfolio value at risk confidence 95% and 99% were 3.111% and 4.400%, the model at the 95% confidence level test results was not significant; 5, from the quality of the model in the sample funds at the Monte Carlo simulation method is better than the other two kinds of methods. Based on the above conclusions, the investment fund or fund company forecast the VaR value distribution hypothesis and empirical distribution trend is consistent in using the analysis method, the GARCH model can not blindly use; fund companies to consider fund risk value calculation of optimal model, enhance the reliability of risk prediction, to strengthen the fund's portfolio risk measurement and monitoring, the risk prevention and management.

【学位授予单位】：上海师范大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F224;F832.51

【参考文献】