A1212
Title: Online detection of risk instabilities based on conditional expectiles
Authors: Matus Maciak - Charles University (Czech Republic) [presenting]
Michal Pesta - Charles University (Czech Republic)
Gabriela Ciuperca - University Lyon-I (France)
Abstract: The purpose is to consider an online changepoint detection procedure based on conditional expectiles, which are convenient tools for risk assessment tasks in empirical finance. The expectiles are well-known in econometrics for being a coherent and elicitable risk measure. In addition, the approach based on the expectiles introduces some robustness when compared with traditional moment-based techniques, and it also provides a more complex insight into the overall data-generating mechanism. The proposed statistical test is proven to be consistent, while the distribution under the null hypothesis does not depend on the underlying functional form or the unknown parameters. Theoretical details and finite sample performance are discussed in the talk.
