微信扫一扫
实时资讯全掌握
|
The limited amount of operational data necessitates extrapolating beyond the sparse data points. Non-parametric methods are not appropriate because the observations provide no guidance beyond the largest realization. Hence, a parametric model is needed, and the choice depends on the underlying assumption of the loss-generating process. Which of the following statements is not an appropriate assumption for generating loss data? A. Multiple loss-generating mechanisms will generally yield similar extreme events. B. A single loss-generating mechanism is assumed for all losses regardless of source. C. There are several underlying loss-generating mechanisms, but some are more likely to yield extreme events. D. The extreme events are not drawn from a known distribution and/or do not offer a pattern for estimation. |