风险度量ES的非参数估计

Nonparametric Estimation of Expected Shortfall

期望损失(Expected Shortfall,ES)是当今最流行的金融资产风险度量的工具之一,是一个理想的一致性风险度量。本文在α-混合序列具有幂衰减混合系数的条件下,讨论了风险度量ES的样本估计量的性质,得到估计量的Bahadur表示以及渐近正态性。

Expected Shortfall is one of the most popular measure of the risk management for financial assets, and is an ideal coherent risk measure. In this paper, we discuss the sample quantile estimation of ES under polynomial decay coefficients of α-mixing time series. We obtain a Bahadur representation and the asymptotic normality of the estimation.