Risk Management and Financial Institution Chapter 12 —— VaR and Es
typora-copy-images-to: Risk Management and Financial Institution
文章目录
- Risk Management and Financial Institution Chapter 12 —— VaR and Es
Risk Management and Financial Institution Chapter 12 —— VaR and Es
- Value at risk (VaR) and expected shortfall (ES) are attempts to provide a single number that summarizes the total risk in a portfolio
12.1 Definition of VaR
- It is the loss level during a time period of length T that we are X% certain will not be exceeded
12.2 Examples (略)
12.3 Drawback of VaR
12.4 Expected Shortfall
-
又称作为conditional value at risk,expected tail loss
-
ES is the expected loss during time T conditional on the loss being greater than the VaR
-
ES has better properties than VaR in that it always recognizes the benefits of diversification , 分散化的收益可以被算入
-
ES的缺陷:
- 比VaR更加难以理解
- 回测过程也更加难
12.5 Coherent Risk Measures
-
风险度量应该满足一定的性态:
- monotonicity 单调性 收益小,意味着风险大
- Translation Invariance 平移不变性 加入K的cash ,总风险减去cash
- Homogeneity 同质性 保持组合中单个资产所占比例不变,调整组合大小,风险相应增减
- Subadditivity 次可加性 两个交易组合组成的新交易组合风险度量小于等于单独的和
-
VaR 并不满足次可加性
-
ES 测量满足四种性态
12.5.1 Exponential Spectral Risk Measures
-
A risk measure can be characterized by the weights it assigns to percentiles of the loss distribution
-
VaR 给了百分之一百的权重在分界点上
-
One idea is to make the weight assigned to the qth percentile proportional to where γ is a constant. This is referred to as the exponential spectral risk measure
-
不同的γ值 ,不同的指数形态
12.6 VaR 和 ES 参数的选择
- 两个参数必须选择,时间区间以及置信区间
- 假设时间界限内,交易组合的变化服从正态分布,X 是置信区间
- ES 的估计公式如下,假设loss is normally distributed
-
Y is the Xth percentile point of the standard normal distribution
-
This shows that when μ is assumed to be zero, ES, like VaR, is proportional to σ
12.6.1 Time Horizon 展望期
- 流动性差的组合,展望期长
- market risks 展望期一般为1天
- 假设:when the changes in the value of the portfolio on successive days have independent identical normal distributions with mean zero
12.6.2 Impact of Autocorrelation 自相关的影响
- 现实中,组合价值的变化并不独立
- 如果天与天之间的组合价值变化并不独立,那么其实用平方根法则预估的VaR值偏低
12.6.3 Confidence Level 置信区间
- 选择置信区间的因素:
- 与信用评级挂钩,AA级99.97%置信区间,一年的展望期
12.7 Marginal, Incremental and Component Measures 边际VaR,递增VaR及成分VaR
-
The marginal value at risk for the ith subportfolio is the sensitivity of VaR to the amount invested in the ith subportfolio交易组合价格变化同某个组合成分变化的比率
-
If an asset’s beta is high, its marginal VaR will tend to be high. If its beta is low, the marginal VaR tends to be low
-
incremental value at risk 是指子交易组合对VaR的递增效应,是组合包含与不包含此子组合VaR的差
12.8 Euler’s Theorem 欧拉定理
- 欧拉定理与风险度量的第三性态相一致
- 当V 为 VaR时:
- 组合VaR 相当于Component VaR的相加
- ES的性质与VaR一致:
- 欧拉定理是用于 risk budgeting 的很好的工具
12.9 Aggregating VaRs and ESs VaR的聚合
- 不同部门的VaR聚合成total VaR
- ρ 代表着损失与损失之间的相关性
- 上式当损失与收益的预期为0时严格成立,其他时候也是很好的估计
12.10 Back- Testing 回测
-
It is a test of how well the current procedure for calculating the measure would have worked in the past 历史的检验
-
VaR 的回测要比ES来得容易,因此监管者在制定市场风险时转换为ES很勉强
-
Back-testing involves looking at how often the loss in a day would have exceeded the one-day 99% VaR when the latter is calculated using the current procedure
-
实际损失超出VaR值叫做exceptions
-
在回测时,我们应该考虑展望期内交易组合本身的变化
- 假定展望期内组合构成没有变化
- 与交易组合真实变化进行比较
-
In practice, risk managers usually compare VaR to both hypothetical portfolio changes and actual portfolio changes
-
Kupiec test (双尾检验)与3.84做比较,大于3.84 则拒绝模型
-
Generally speaking, the difficulty of back-testing a VaR model increases as the VaR confidence level increases
12.10.1 Bunching 聚束效应
- 使用Chiristofferson test 例外的发生是均匀分布的,还是聚束分布的