波动性变化下的VaR历史模拟法实证研究

VaR(在险价值)方法是当今运用得最为广泛的金融市场风险度量方法。历史模拟法作为计算VaR的主要方法之一,其计算出来的VaR的风险度量效果需要得到现实金融市场数据的检验。本文通过选取上证综指日收益率的历史数据,分别在市场波动性不发生改变、市场波动性变大和市场波动性变小三种情况下对历史模拟法的有效性进行检验,检验结果表明在市场波动性不发生变化的情况下,历史模拟法计算的VaR能够有效地度量市场风险;在市场波动性变大的情况下,该方法会低估市场风险;在市场波动性变小的情况下,该方法会高估市场风险。通过对历史模拟法的计算原理进行分析,历史样本自身的时滞性是导致该方法在市场波动性发生变化时无效的原因。

An Empirical Study of Historical Simulation Method of VaR with Changes in Volatility

VaR(value at risk)method is the most popular tool in risk measure of financial market now. Historical simulation method is one of the main methods that calculate VaR, and its effectiveness needs to be tested by actual data from financial market. This paper selects the historical data of daily yield rate of the Shanghai Composite Index, and empirically tests the effectiveness of historical simulation method under three kinds of volatility changing conditions. The empirical result shows that VaR calculated by historical simulation method will measure market risk effectively when the volatility of market does not change. However, the method will underestimate the risk when the volatility of market climbs, and overestimate the risk when the volatility of market declines. By analyzing its calculation principles, we find the hysteresis of historical data. This characteristic is the reason why the method is ineffective when volatility of market changes. So, investors who use it to measure market risk must consider changes of volatility in the future.