tl
Traffic light test for value-at-risk (VaR) backtesting
Syntax
Description
Examples
Input Arguments
Output Arguments
More About
算法
交通灯测试基于二项式分布。认为N
是观察的数量,p
= 1 -VARLEVEL
is the probability of observing a failure if the model is correct, andxis the number of failures.
The test computes the cumulative probability of observing up toxfailures, reported in the'可能性'
column,
where 是带有参数的二项式变量的累积分布N和p, withp= 1 -VARLEVEL。这三个区域是根据此累积概率定义的:
Green: ≤
0.95
Yellow:
0.95
< ≤0.9999
红色的:
0.9999
<
The probability of a Type-I error, reported in the'typei'
列,是
。
如果模型正确,则这种概率对应于错误拒绝模型的概率。Probability和TypeI不要总结1,它们超过1,恰好有可能x失败。
The increase in scaling factor, reported in the'增加'
列,总是0
为了green
zone and always1
为了red
zone. For theyellow
zone, it is an adjustment based on the relative difference between the assumed VaR confidence level (VARLEVEL)和观察到的置信度(x/N), whereN
是观察的数量和xis the number of failures. To find the increase under the assumption of a normal distribution, compute the critical valuesZassumed和zObserved。
The increase to the baseline scaling factor is given by
with the restriction that the increase cannot be negative or greater than1
。The baseline scaling factor in the Basel rules is 3.
Thetl
function computes the scaling factor following this methodology, which is also described in the Basel document (see参考). Thetl
function does not apply any ad-hoc adjustments.
参考
[1]巴塞尔银行监督委员会,使用“回测”的监管框架以及内部模型的市场风险资本需求。January, 1996,https://www.bis.org/publ/bcbs22.htm。