On the Use of Distributional Methods in Ratemaking. Reply
回应了Venezian先生对费率制定中分布方法使用的质疑，指出实践中费率制定主要未采用理论统计分布，但存在例外（如ISO对商业险超额限额的定价），并强调作者模型的有效性未受影响，后续将提交实证证据。
As Mr. Venezian correctly suspects, the statement he quotes is intended to describe ratemaking only in the narrow sense, that is, the way it is generally done in practice. The authors are aware of the procedure followed by the Insurance Service Office for pricing excesss limits in commercial lines, but this is an isolated case representing the exception to the rule. Whether taken in terms of rules and regulations promulgated by the states, or in terms of practices followed by insurance companies in filing for rate revision in the various lines of insurance, the rating process fails, in the main, to make use of theoretical statistical distributions. The authors had believed that the use of the modifying word basically in the statement referred to by Mr. Venezian would be sufficient to remind the attentive reader of the existence of some exceptions though minor in nature. If this was not adequate, then we regret not including a footnote citing as an example the ISO exception which would have alleviated any possibility of a false impression however peripheral in nature. Neither the validity of the proposed ratemaking model, nor the reader's understanding and appreciation of it, has been compromised. The authors have since gathered strong empirical evidence in support of their model and intend to submit their findings soon to this Journal.