Based on Ernst & Young's experience in dealing with actuarial departments for major life insurance companies, it was not a huge surprise the professional services organization found significant data quality issues in its report "From the Actuarial Transformation Roundtable," according to Steve Goren, a senior manager in the actuarial advisory service group for Ernst & Young. "There are few or no strong controls in place over data quality and data management processes," he says. Mike Hughes, a partner with Ernst & Young, believes the fundamental message derived from the insurers taking part in the roundtable is the business processes are struggling from a lack of effective data management strategies for the actuarial departments. "Insurers don't have the right sort of database management systems and actuarial data warehouses in place to deal with some of the business needs," he says.

One of the challenges companies face–and a reason they struggle with managing data–is the source data they need for the different actuarial processes resides in multiple legacy administration systems, commission systems, reinsurance systems, and accounting systems. Many of those different source systems, Hughes asserts, have data quality issues that need to be addressed. "If you don't have a good data management system in place and the right tools and infrastructure, getting data from point A to point B can be a real business challenge," he says.

Data governance can't be limited to a line of business, points out Goren. It needs active participation with each of the lines of business as well as the system owners–whether IT or the businesses. "Data is consumed from a number of different systems, and the systems are used for different purposes, so it needs to be an integrated approach to manage the data across all the different data providers and consumers," he says. "It can't be just an IT solution or a single line of business solution. It can, however, begin within a line of business in getting the data quality issues identified and then pass those issues up to the source systems that provide the data."