【摘要】：在刻畫(huà)和估計(jì)資產(chǎn)聯(lián)合損失分布函數(shù)的基礎(chǔ)上對(duì)開(kāi)放式基金VaR風(fēng)險(xiǎn)進(jìn)行比較分析。在極值情況下有兩種算法:一種是半?yún)?shù)法,采用GPD分布來(lái)擬合損失分布的尾部和核密度分布擬合損失分布的中間部分,運(yùn)用copula函數(shù)來(lái)刻畫(huà)資產(chǎn)損失的相依結(jié)構(gòu);另一種是非參數(shù)法,用Bootstrapping和FHS方法對(duì)收益率進(jìn)行抽樣和模擬。經(jīng)過(guò)實(shí)證分析發(fā)現(xiàn)在較低置信水平下,宜于采用非參數(shù)法;而在較高置信水平下,采用半?yún)?shù)法則更合適,這也充分說(shuō)明半?yún)?shù)法適合在更極值情形下對(duì)開(kāi)放式基金估計(jì)VaR風(fēng)險(xiǎn)。
[Abstract]:On the basis of characterizing and estimating the joint loss distribution function of assets, this paper makes a comparative analysis on the VaR risk of open-end funds. There are two algorithms under extreme value: one is semi-parametric method, which uses GPD distribution to fit the tail of loss distribution and kernel density distribution to fit the middle part of loss distribution, and uses copula function to describe the dependent structure of asset loss; The other is non-parametric method, which uses Bootstrapping and FHS to sample and simulate the return rate. Through the empirical analysis, it is found that the nonparametric method is suitable for the lower confidence level. In the case of higher confidence level, it is more appropriate to adopt the semi-parametric rule, which fully shows that the semi-parametric method is suitable for estimating the VaR risk of open-end funds under the condition of more extreme value.
【作者單位】： 四川大學(xué)經(jīng)濟(jì)學(xué)院;華僑大學(xué)經(jīng)濟(jì)與金融學(xué)院;
【基金】：教育部人文社科基金資助項(xiàng)目(09XJC90007)
【分類號(hào)】：F224;F830.9

【參考文獻(xiàn)】

相關(guān)期刊論文 前1條

1 陳小紅;何浩;;開(kāi)放式基金的VaR值測(cè)算與評(píng)估——基于GARCH模型的實(shí)證分析[J];武漢金融;2006年08期

【共引文獻(xiàn)】

相關(guān)碩士學(xué)位論文 前5條

1 朱s搕，

本文編號(hào)：2346264