流动性调整地风险价值度量:基于金融高频数据的实证分析  

Liquidity adjusted VaR measurement based on high frequency data

作　　者：刘晓星[1,2] 

机构地区：[1]广东商学院金融学院,广州510320 [2]复旦大学金融研究院,上海200433

出　　处：《系统工程理论与实践》2009年第7期16-26,共11页Systems Engineering-Theory & Practice

基　　金：中国博士后基金(20070410665);教育部人文社科项目(07JC790022);广东省自然科学基金(7301175)

摘　　要：结合ACD和UHF-GARCH理论构造了适应我国股票市场的日内风险价值WACD(1,1) -UHF-GARCH(1,1)-IVaR模型,然后引入价格冲击模型构造的流动性度量指标,以浦发银行为例对我国股票市场的流动性日内风险价值进行了实证分析,得出三点结论:一是我国股票市场的交易持续期具有很强的聚类性;二是高频数据中GARCH效应显著,且利好消息会对等量的利空消息产生更大的波动,但考虑流动性的影响因素后利空利好消息对市场的冲击效果都有明显降低;三是蒙特卡罗模拟结果表明,不考虑流动性影响的风险价值较大幅度地低估了实际损失。This paper constructed a WACD（1,1）-UHF-GARCH（1,1）-IVaR Model to the stock market of China with the theory of ACD and UHF-GARCH, and take Shanghai Pudong Development Bank as an example, gave an empirical analysis to the intraday Value at Risk adjusted by liquidity in our stock market combining the liquidity index designed by price impact model. The result shows that： firstly, the transactions duration has strong clustering character;secondly, there also has strong CARCH effect in high-frequency data, and good news will come out more volatility than bad news, but the effect of both news impacting on the market have obviously reduced after considering the influence of liquidity; thirdly, Monte Carlo simulation shows that the Value at Risk will underestimate the real losing without considering the liquidity effects.

关 键 词：高频数据 WACD-UHF-GARCH-IVaR 蒙特卡罗模拟 流动性调整的VaR 

分 类 号：F830.91[经济管理—金融学]