当前位置：中博教育 > FRM^® > 学习指导 > FRM^®知识点：Backtesting VaR Models

FRM^®知识点：Backtesting VaR Models

文章来源：网络

发布时间：2022-01-25 15:15

阅读：516次

本文为大家讲解FRM®知识点：var模型的回测，（Backtesting VaR Models），一起来学习吧。

As with any type of modelling,

a VaR model must be validated.In particular,

backtesting has been the industry

standard for validating VaR models.

修改后_9.jpg

Backtesting Approaches：

Banks typically draw inference on the performance of VaR models using backtesting exceptions.

银行通常会利用回测来推断VaR模型的是否正确。

For regulatory capital,the MRA imposes a multiplier on VaR depending on the number of backtesting exceptions the bank experiences.【点击免费下载>>>更多FRM学习相关资料】

In addition,the choice of whether to evaluate a VaR model on the basis of hypothetical or actual backtesting may be affected by the characteristics of the portfolio.

假设或实际的回测是否会受到投资组合特征的影响。

Indeed,actual backtesting is less informative when the composition of the portfolio has reccently changed

实际上，当投资组合的构成最近发生变化时，实际的回溯测试提供的信息较少

it may impose substantial computational burdens because it requires reconstructing the history of the portfolio on the basis of its current composition.

假设的回测可能会带来大量的计算负担，因为它需要在当前组合的基础上重建投资组合的历史。

“Kupiec(1995)introduced the unconditional coverage likehood ratio tests as inference tools for whether the VaR model generated the correct number of exceptions”——————引入无条件覆盖

“Christoffersen(1998)has proposed a conditional backtesting exception test that accounts for the timing as well as the number of exceptions”

——————提出了一个有条件的回溯异常测试，该测试考虑了时间和异常的数量。

“Pritsker(2006)and Shang(2009),also consider the use of Mean-Squared-Error(MSE)as a measure of VaR performance in backtesting”

——————————————在回测中考虑使用MSE

“Lopez(1999),for instance,formalised this idea by introducing a quadratic loss function where loss is the difference between actual P＆L and VaR,when an exception occurs.”————引入一个二次损失函数

2022年FRM^®最新学习资料包

请大家认真填写以下信息，获取2025年FRM^®学习资料包，会以网盘链接的形式给到大家，点击免费领取后请尽快保存。