微信扫一扫
实时资讯全掌握
|
Which of the following cases leads to the least accurate confidence interval for VAR estimates? A. The investigator knows that the data is Weibull-distributed so a confidence interval is not needed, as this is the true and exact, non-stochastic value. B. The investigator knows that the return distribution is Weibull-distributed and can construct a 95% confidence interval by observing the lowest 2.5% simulated observation for the lower bound, and the upper 97.5% observation for the upper bound. C. The investigator knows that the true return distribution is Weibull-distributed; however, the investigator uses a normal quantile function as a best alternative to estimate VAR. D. The investigator assumes a normal returns distribution and does not know or ignores that the true data is Weibull-distributed and uses sample moments and a normal quantile function to estimate VAR. |