Forecasting Value-at-Risk and Expected ShortfallUsing Range CARE Models
| Autor: | Chien-Yu Shen, 沈謙昱 |
|---|---|
| Rok vydání: | 2012 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 100 This thesis considers the conditional autoregressive expectiles (CARE) with the intra-day high-low price range. Inference, quantile forecasting and model comparison for CARE is investigated. A Bayesian method to forecast Value-at-Risk (VaR) and Expected Shortfall (ES) is employed and an adaptive Markov chain Monte Carlo scheme is designed. The proposed methods are illustrated using ve inter- national stock market return series. Three backtesting tests are used to assess the VaR and ES forecasting performance. The nonlinear CARE models are found to exhibit superior performance. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|