Change-point detection for expected shortfall in time series |
| |
Affiliation: | 1. Department of Mathematical Sciences, Tsinghua University, Beijing 100084, China;2. Center for Statistical Science, Department of Industrial Engineering, Tsinghua University, Bejing 100084, China |
| |
Abstract: | Expected shortfall (ES) is a popular risk measure and plays an important role in risk and portfolio management. Recently, change-point detection of risk measures has been attracting much attention in finance. Based on the self-normalized CUSUM statistic in Fan, Glynn and Pelger (2018) and the Wild Binary Segmentation (WBS) algorithm in Fryzlewicz (2014), this paper proposes a variant WBS procedure to detect and estimate change points of ES in time series. The strengthened Schwarz information criterion is also introduced to determine the number of change points. Monte Carlo simulation studies are conducted to assess the finite-sample performance of our variant WBS procedure about ES in time series. An empirical application is given to illustrate the usefulness of our procedure. |
| |
Keywords: | Change point Expected shortfall Risk measure C12 C22 C32 |
本文献已被 ScienceDirect 等数据库收录! |
|