论文部分内容阅读
本文研究了金融风险管理理论中风险价值(VaR)的非参数核光滑估计和经验估计的效率问题.对非独立的时间序列损失/收益样本,在均方误差(MSE)准则的意义下引入“亏量”的概念,亏量越大表明估计效率越低.并利用亏量对VaR模型的核光滑估计和基于样本分位数的经验估计进行了比较,在理论上证明了VaR模型的核光滑估计优于经验估计.同时,通过计算机模拟证实了理论获得的结论.本文还对国内沪深两市上的证券投资基金进行了实证分析,计算了样本基金的VaR风险度量的经验估计和核光滑估计,并计算了样本基金基于周收益率和VaR估计的风险调整收益(RAROC)值,以此对样本基金的业绩做出了有用的评价.
This paper studies the problem of non-parametric kernel smoothing and empirical estimation of VaR in financial risk management theory.In the sense of mean square error (MSE), non-independent time series loss / The concept of “deficit ” indicates that the larger the deficit, the lower the estimation efficiency, and the comparison between the nuclear smoothing of the VaR model and the empirical estimation based on the sample quantile by using the deficit shows that the VaR model Nuclear smoothing is superior to the empirical estimation.At the same time, the conclusion obtained by the theory is confirmed by computer simulation.This paper also empirically analyzes the securities investment funds in the domestic Shanghai and Shenzhen stock markets, calculates the empirical estimation of the VaR risk measurement of the sample funds and The calculation of the sample fund’s RAROC based on the weekly rate of return and the VaR estimate was made to make a useful assessment of the performance of the sample fund.