Backtesting Parametric Value-at-Risk With Estimation Risk

Escanciano, Juan Carlos; Olmo, José

doi:10.2139/ssrn.975104

Cited by 37 publications

(81 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our tests strongly reject this model for the CAC and FTSE stocks, and it is dubious for the DAX index, with rejections at 10% when d = 2 with CvM n;2 and at 5% when d = 1 with both, CvM n;1 and KS n;1 : Second, the cumulative conditional backtest has rather low power and indeed, it is not able to detect any of these alternatives. This result is consistent with other …nite sample studies using this test, see Escanciano and Olmo (2008). Third, it is apparent from the results for CAC that in order to detect this alternative it is important to consider a larger information set containing the second lag.…”

Section: Application To Market Risk Managementsupporting

confidence: 90%

See 1 more Smart Citation

Specification Tests of Parametric Dynamic Conditional Quantiles

Escanciano

Velasco

2008

SSRN Journal

Self Cite

View full text Add to dashboard Cite

This article proposes omnibus speci…cation tests of parametric dynamic quantile regression models. Contrary to the existing procedures, we allow for a ‡exible and general speci…cation framework where a possibly continuum of quantiles are simultaneously speci…ed. This is the case for many econometric applications for both time series and cross section data which require a global diagnostic tool. We study the asymptotic distribution of the test statistics under fairly weak conditions on the serial dependence in the underlying data generating process. It turns out that the asymptotic null distribution depends on the data generating process and the hypothesized model. We propose a subsampling procedure for approximating the asymptotic critical values of the tests. An appealing property of the proposed tests is that they do not require estimation of the non-parametric (conditional) sparsity function. A Monte Carlo study compares the proposed tests and shows that the asymptotic results provide good approximations for small sample sizes. Finally, an application to some European stock indexes provides evidence that our methodology is a powerful and ‡exible alternative to standard backtesting procedures in evaluating market risk by using information from a range of quantiles in the lower tail of returns.

show abstract

Section: Application To Market Risk Managementsupporting

confidence: 90%

“…Kupiec, 1995), see also Christo¤ersen (1998) and Escanciano and Olmo (2008), based on the absolute value of the standardized sample mean, i.e.…”

Section: Application To Market Risk Managementmentioning

confidence: 99%

Specification Tests of Parametric Dynamic Conditional Quantiles

Escanciano

Velasco

2008

SSRN Journal

Self Cite

View full text Add to dashboard Cite

show abstract

“…That is, we have (i) used only a very parsimonious number of parameters, (ii) focused on normally distributed errors and (iii) specified a standard GARCH equation of the Bollerslev (1986) type. However, as highlighted by Escanciano and Olmo (2010), model misspecification can seriously bias VaR estimators and related procedures. Therefore, to shed some light on the impact of different (potentially more adequate) filter settings on our results, we perform three robustness checks.…”

Section: General Specification Testsmentioning

confidence: 99%

“…35 We have also implemented the bootstrap-based procedures of Escanciano and Olmo (2010) which are designed to address the issue that classic coverage tests are affected by model misspecification in conditional VaR models.…”

Section: Crisis Subsamplesmentioning

confidence: 99%

How accurate are modern Value-at-Risk estimators derived from extreme value theory?

Mögel

Auer

2017

Rev Quant Finan Acc

View full text Add to dashboard Cite

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. In this study, we compare the out-of-sample forecasting performance of several modern Valueat-Risk (VaR) estimators derived from extreme value theory (EVT). Specifically, in a multiasset study covering 30 years of stock, bond, commodity and currency market data, we analyse the accuracy of the classic generalised Pareto peak over threshold approach and three recently proposed methods based on the Box-Cox transformation, L-moment estimation and the Johnson system of distributions. We find that, in their unconditional form, some of the estimators are acceptable under current regulatory assessment rules but none of them can continuously pass more advanced tests of forecasting accuracy. In their conditional forms, forecasting power is significantly increased and the Box-Cox method proves to be the most promising estimator. However, it is also important to stress that the traditional historical simulation approach, which is currently the most frequently used VaR estimator in commercial banks, can not only keep up with the EVT-based methods but occasionally even outperforms them (depending on the setting: unconditional vs. conditional). Thus, recent claims to generally replace this simple method by theoretically more advanced EVT-based methods may be premature. Terms of use: Documents inJEL-Codes: G100, G110, G170.

show abstract

“…The proposed backtesting methodology is based on the multivariate tests developed by Chitturi (1974) and Hosking (1980). However, since we apply the multivariate portmanteau test to multinomial, non-continuous-type data, we derive its limit distribution under the null hypothesis following the approach of Escanciano and Olmo (2010). Escanciano and Olmo (2010) showed that tests which do not consider the impact of estimation risk may use wrong critical values to assess market risk.…”

Section: Introductionmentioning

confidence: 99%

Backtesting Aggregate Risk

Danciulescu

2015

Journal of Forecasting

View full text Add to dashboard Cite

This paper proposes the implementation of a VaR backtesting procedure able to overcome the subadditivity property failure of value‐at‐risk (VaR). More precisely, we propose the implementation of a multivariate portmanteau test statistic of Ljung–Box type applied to hits collected from several trading desks or divisions at once. Simulation \exercises illustrate that this method is testing for aggregate risk, accurately accounting both for diversification (negative hit cross‐correlation) and contagion/risk spillovers (positive hit cross‐correlation). An application using profit and loss and VaR data collected for two international major banks illustrates how our proposed backtesting procedure performs in a realistic environment. Copyright © 2015 John Wiley & Sons, Ltd.

show abstract

Backtesting Parametric Value-at-Risk With Estimation Risk

Cited by 37 publications

References 30 publications

Specification Tests of Parametric Dynamic Conditional Quantiles

Specification Tests of Parametric Dynamic Conditional Quantiles

How accurate are modern Value-at-Risk estimators derived from extreme value theory?

Backtesting Aggregate Risk

Contact Info

Product

Resources

About