{"title":"EVT �? CAViaR 모형�? �?�한 국고채 VaR 예측모형 분�? (VaR Forecasts for Korea Treasury Bonds via EVT and CAViaR Models)","authors":"Jaeho Yun","doi":"10.2139/ssrn.3204997","DOIUrl":null,"url":null,"abstract":"Korean Abstract: 본 논문�?� 시장리스�?��?� 주�?� 측정수단�?� VaR(value at risk)�?� 여러 예측모형�?� 한국�?� 국고채(1년, 5년, 10년 만기 할�?�채) 보유수�?�률�? �?용하여 비�? 분�?하였다. 특히 EVT(extreme value theory)를 �?용한 모형과 CAViaR(conditional autoregressive value at risk) 모형�?� 예측성과�? 초�?�?� 맞추었다. 실�? 분�? 결과, 채권 보유수�?�률�?� GARCH(generalized autoregressive conditional heteroskedasticity) 모형�? 필터�?한 후 EVT를 �?용한 EVT-GARCH 모형�?� 성과가 1% �? 5% VaR 예측 모�?�?게서 우수한 것으로 나타나, 기존�? �?리 사용�?�고 있는 역사�? 시뮬레�?�션�?�나 RiskMetrics 등�?� �?�함한 본 논문�?� 고려하고 있는 여러 모형 중�?서 가장 우수한 VaR 예측성과를 나타났다. 한편, CAViaR 모형들�?� 경우 �?�부 모형�?� 1% VaR�?서 우수한 예측성과를 보였지만 EVT-GARCH 모형�?� 성과�?는 미치지 못하였다. English Abstract: In this paper, we compare various forecasting models for VaR(value at risk), which is the main measure of market risk, by applying them to Korea Treasury Bonds (1-year, 5-year, and 10-year maturity discount bonds). In particular, we focus on the predictive performance of the extreme value theory(EVT) models and the conditional autoregressive value at risk(CAViaR) models. Our empirical analysis shows that the EVT-GARCH model, which uses the EVT after filtering bond returns through the GARCH(generalized autoregressive conditional heteroskedasticity) model, performs well in the 1% and 5% VaR forecasts. On the other hand, some CAViaR models show good forecasting performance at 1% VaR, but not sufficient to outperform the EVT-GARCH model.","PeriodicalId":166226,"journal":{"name":"Korea Deposit Insurance Corporation (KDIC) Research Paper Series","volume":"19 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Korea Deposit Insurance Corporation (KDIC) Research Paper Series","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.2139/ssrn.3204997","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Korean Abstract: 본 논문�?� 시장리스�?��?� 주�?� 측정수단�?� VaR(value at risk)�?� 여러 예측모형�?� 한국�?� 국고채(1년, 5년, 10년 만기 할�?�채) 보유수�?�률�? �?용하여 비�? 분�?하였다. 특히 EVT(extreme value theory)를 �?용한 모형과 CAViaR(conditional autoregressive value at risk) 모형�?� 예측성과�? 초�?�?� 맞추었다. 실�? 분�? 결과, 채권 보유수�?�률�?� GARCH(generalized autoregressive conditional heteroskedasticity) 모형�? 필터�?한 후 EVT를 �?용한 EVT-GARCH 모형�?� 성과가 1% �? 5% VaR 예측 모�?�?게서 우수한 것으로 나타나, 기존�? �?리 사용�?�고 있는 역사�? 시뮬레�?�션�?�나 RiskMetrics 등�?� �?�함한 본 논문�?� 고려하고 있는 여러 모형 중�?서 가장 우수한 VaR 예측성과를 나타났다. 한편, CAViaR 모형들�?� 경우 �?�부 모형�?� 1% VaR�?서 우수한 예측성과를 보였지만 EVT-GARCH 모형�?� 성과�?는 미치지 못하였다. English Abstract: In this paper, we compare various forecasting models for VaR(value at risk), which is the main measure of market risk, by applying them to Korea Treasury Bonds (1-year, 5-year, and 10-year maturity discount bonds). In particular, we focus on the predictive performance of the extreme value theory(EVT) models and the conditional autoregressive value at risk(CAViaR) models. Our empirical analysis shows that the EVT-GARCH model, which uses the EVT after filtering bond returns through the GARCH(generalized autoregressive conditional heteroskedasticity) model, performs well in the 1% and 5% VaR forecasts. On the other hand, some CAViaR models show good forecasting performance at 1% VaR, but not sufficient to outperform the EVT-GARCH model.
evt�?富裕模型�?�?是国库债VaR预测模型吗?(VaR Forecasts for Korea Treasury Bonds via EVT and CAViaR Models)
Korean Abstract:本论文?我知道了,我知道了��?是测量手段吗?�var (value at risk)�?是各种预测模型吗?�韩国�?国库债(1年、5年、10年满期)持有数呢?�?还真灵验生气了。特别是EVT(extreme value theory) ?灵活的模型和CAViaR(conditional autoregressive value at risk)模型这是预测的成果初��?凑齐了。�?分�?结果债券持有数是多少?这是generalized autoregressive conditional heteroskedasticity模型过滤器�?之后还会做EVT ?这是神准的EVT-GARCH模型成果是1%吗?百分之五的虚拟实技预测,什么?在这里表现得很优秀啊�?使用�?今天的历史是什么?是方程式吗?是方程式吗?还有风险投资等等��?这篇论文还算不错你在考虑的各种模型中西显示出最优秀的VaR预测成果。另一方面,你知道那些有线电视模型吗?��?是模型吗?�1% var�?虽然取得了优秀的预测成果,但这是EVT-GARCH模型��成果?——编者注】没有达到。english abstract:In this paper, we compare various forecasting models for VaR(value at risk), which is the main measure of market risk, by applying them to Korea Treasury Bonds (1-year, 5-year,10-year maturity discount bonds。the predictive performance of the extreme value theory(EVT) models and the conditional autoregressive value at risk(CAViaR) modelsOur empirical analysis shows that the EVT-GARCH model, which uses the EVT after filtering bond returns through the GARCH(generalized autoregressive conditional heteroskedasticity) modelperforms well in the 1% and 5% VaR forecasts。On the other hand, some CAViaR models show good forecasting performance at 1% VaR, but not sufficient to outperform the EVT-GARCH model。