论预测性信息行政责任与民事责任的认定差异

On the Differences in the Determination of Administrative Liability and Civil Liability Regarding Predictive Information

摘要: 实践中，同一预测性信息可能在行政监管与民事司法中获得不同法律评价，反映出预测性信息行政责任与民事责任认定标准的差异。通过对典型案件的分析可以发现，行政审查通常以结果偏差为审查起点，并重点考察预测形成过程中的审慎义务；民事司法则以行为发生时点为判断基准，以编制基础是否“明显不合理”等作为重要判断标准。两种责任在制度功能与评价标准上的分化具有一定合理性，但认定标准长期分离也会增加市场主体的合规不确定性，并影响发行人及中介机构的行为选择。因此，应以“合理基础”为共同判断纽带，在统一基础行为评价的同时，根据行政责任与民事责任的不同制度功能构建分层责任体系。

Abstract: In practice, the same predictive information may be subject to divergent legal assessments under administrative regulation and civil litigation, reflecting differences in the criteria for determining administrative and civil liability regarding such predictive information. An analysis of typical cases reveals that administrative review typically begins with assessing outcome deviation and places particular emphasis on the duty of prudence exercised during the formation of the prediction; in contrast, civil litigation adopts the moment of the conduct as its benchmark for judgment, employing criteria such as whether the underlying basis for the prediction was “manifestly unreasonable” as key evaluative standards. While the divergence in institutional functions and evaluation criteria between these two types of liability is somewhat justified, the prolonged separation of these assessment standards may increase compliance uncertainty for market participants, thereby influencing the behavioral choices of issuers and intermediaries. Consequently, it is advisable to adopt the “reasonable basis” as a common cornerstone for evaluation, thereby unifying the assessment of conduct while establishing a tiered liability framework tailored to the distinct institutional functions of administrative and civil liability regimes.