我国金融机构操作风险的特征分析与防控——基于1987—2012年媒体公开报道事件数据

Characteristics of Operational Losses Occurred in Chinese Financial Institutions: An Empirical InvestigationBased on Event Data from Public Media Sources over 1987-2012

操作风险涵盖的细分风险类别很广,其中就包含有会直接导致金融机构倒闭的低频高危型风险。因此,如何精确估算出金融机构为操作风险所应预留的经济资本额是学界和业界共同关注的问题。该问题的解决依赖于大量的内外部数据相结合,内部数据可由金融机构自我统计,但来自媒体公开报道的外部数据不足却一直是操作风险研究人员的一大困扰。为增加我国金融机构操作风险的外部数据样本,本文在总结操作风险特征的基础上规范了外部数据的搜集准则,并根据这套准则整理出了3336条发生于1987—2012年间的操作风险事件。利用该数据库,本文不仅分地区、分机构性质和机构级别测算了操作风险损失的次数和强度,还对中外操作风险损失各自的特征进行了比较。结果表明:我国金融机构的操作风险多发于经济发达地区、基层金融企业、管理相对薄弱的机构以及组织架构复杂的大型商业银行,对此需要加强防控。

Operational risk encompasses a variety of second--tier risk types, including the low一frequency high一severity type of fat tailed risk that would directly lead to the bankruptcy of financial institutions. As a result, how to accurately estimate the amount of economic capital required ex一ante for operational risk exposures has become a key question of both industry's and academic interests. The solution to this problem relies on the combi nation of large internal and external datasets. While internal data can be collected by financial institutions themselves, the external data from public media sources are limited due to many reasons, especially for Chinese financial institutions. To expand the sample size of external operational loss data, this paper has constructed a database with 3,336 observations over the period from 1987一2012. Using this database, our paper computes the average frequency and severity of Chinese operational losses for different geographical districts, firm types, and levels of branches. We further compare the Chinese dataset with global operational risk data. The findings are that operational losses in Chinese Financial institutions occurred more often in provinces with higher GDP per capita, firm branches of lower hierarchy and with weaker administration skills, and larger commercial banks with a higher degree of complexity in terms of organizational structure.