NO.PZ2016070202000015
问题如下：
Brenda Williams is a risk analyst who wants to model the dependence between asset returns using copulas and must convince her manager that this is the best approach. Which of the following statements are correct?
I. The dependence between the return distributions of portfolio assets is critical for risk measurement.
II. Correlation estimates often appear stable in periods of low market volatility and then become volatile in stressed market conditions. Risk measures calculated using correlations estimated over long horizons will therefore underestimate risk in stressed periods.
III. Pearson correlation (ρ) is a linear measure of dependence between the return of two assets equal to the ratio of the covariance of the asset returns to the product of their volatilities.
IV. Using copulas, one can construct joint return distribution functions from marginal distribution functions in a way that allows for more general types of dependence structure of the asset returns.
选项：
A.I, II, and III
B.II and IV
C.I, II, III, and IV
D.I, III, and IV
解释：
D is correct. The dependence is critical, so statement I. is correct. The usual Pearson correlation is a linear measure of dependence, so statement III. is correct. Statement IV. is also correct. For statement II., correlations indeed change over stressed periods, but it is not clear whether this biases long-term correlations upward or downward. Also, the effect on the portfolio risk depends on the positioning. Hence, there is not enough information to support statement II.
以下个人理解，老师看看是否对：
1.正常时期波动小，压力时候波动大，但是波动大可能导致ρ往下走得多，也可能导致ρ往上走得多。
2.正常时期ρ小，压力时期ρ大，这个只是大部分情况来说是这样，特殊情况有可能压力时期ρ减小？
3.考虑ρ变大或者变小，还要考虑头寸的问题，例如全部投了equity，那么ρ变大反而是好的，风险会更小？
所以综上所述，statement II就是错的，对吗？
