Abstract:Predicting the occurrence of tail events is of great importance in financial risk management. By employing the method of peak-over-threshold (POT) in extreme value theory (EVT) to identify the financial extremes, we perform a recurrence interval analysis (RIA) on these extremes. We find that the waiting time between consecutive extremes (recurrence interval) follows a q-exponential distribution, and the sizes of extremes above the thresholds (exceeding size) conform to a generalized Pareto distribution. We also find that there is a significant correlation between recurrence intervals and exceeding sizes. We thus model the joint distribution of recurrence intervals and exceeding sizes through connecting the two corresponding marginal distributions with the Frank and Ali-MikhailHaq (AMH) copula functions, and apply this joint distribution to estimate the hazard probability to observe another extreme in increment t time since the last extreme happened t time ago. Furthermore, an extreme predicting model based on RIA-EVT-Copula is proposed by applying a decision-making algorithm on the hazard probability. Both in sample and out-of-sample tests reveal that this new extreme forecasting framework has better predicting performance than the forecasting model based on the hazard probability only estimated from the distribution of recurrence intervals. Our results not only shed a new light on understanding the occurring pattern of extremes in financial markets, but also improve the accuracy of predicting financial extremes for risk management.(C) 2022 Elsevier B.V. All rights reserved.

The Financial Threshold's Improvement of Hill Estimation——Based on An Empirical Analysis of POT Model in EVT

Improving the traditional Hill estimator to select the threshold in extreme value theory

A New Method to Choose the Threshold in the Pot Model

A New Method to Choose the Threshold in Extreme Value Distribution: Applying the Change Point Theory

Evaluation of Conditional VaR Based on Threshold Quantile Regression Model

Quantification of Statistical Uncertainties in Performing the Peak over Threshold Method

The Estimating Method of VaR Based on the Threshold Double AR Model and Its Application

Assessing the importance of the choice threshold in quantifying market risk under the POT approach (EVT)

2T-POT Hawkes model for left- and right-tail conditional quantile forecasts of financial log-returns: out-of-sample comparison of conditional EVT models

Measuring Financial Risks by Peak over Threshold Method

Changing Rule and Influence of Threshold for Peak-over-threshold Flood Series under the Changing Environment in Wujiang, China

Predicting Tail Events in a RIA-EVT-Copula Framework

One Method of Application of Extreme Value Theory to Computing Value-at-Risk

Increasing the risk management effectiveness from higher accuracy: A novel non-parametric method

Estimation of Extreme Value-at-Risk: an EVT Approach for Quantile GARCH Model

Evaluation of Peak over Threshold Approach for Road Safety Estimation

Estimation of CVaR and Empirical Analysis Based on Tail Index Regression Model

Efficient Estimation in Extreme Value Regression Models of Hedge Fund Tail Risks

Liquidity Risk Measurement of Commercial Bank Based on EGARCH-POT Model

The Financial Risk Measurement EVaR Based on DTARCH Models

The Threshold GARCH Model: Estimation and Density Forecasting for Financial Returns*