Print the Legend

“When the legend becomes fact, print the legend.” – The Man Who Shot Liberty Valance

Not too long ago, I passed a rainy weekend watching old westerns, and I was struck by the relevancy of the above line during a recent discussion with an insurance client about legacy administration system conversions.

We were talking about the approaches to migrating old blocks of policies from first-gen policy admin systems (PAS) to a new system. This usually entails mapping all data and transactions associated with 20- to 30-year-old policies including historical values. Since the systems need to allow for the reversal and reapplication of premiums, loans, withdrawals, and all other forms of financial activity, many companies adopt an approach that sets up each policy at a specific point in the past (usually two years) and then reprocesses the transactions on the new system. The resulting values on the new system are compared against the old system at the same point in time.

A large percentage of the policies (more than 80%) will usually be in balance within a small tolerance of error. Then the fun begins—the conversion team analyzes the various reasons behind the differences. These fall into the following categories:

• Product rules incorrectly configured in the new system, or it does not support specific features
• Differences in rounding and/or order of processing rules between old and new system
• Missing transactions from the old system
• Product rules were incorrectly configured on the old system
• System errors/bugs exist in the old system
• Future date processing not coded in the old system
• Hard coding of rates and values to correct system errors or make policyowners whole

Dealing With Differences

Many of these differences can be addressed by taking action on the new system. Others are addressed by cleaning up the data from the old system. But these methods of handling differences assume that the old system contains the correct values.

However, a good deal of discovery relates to the values being incorrect in the old system, and they may have been incorrect for many years. These values were communicated to the policyowner each year in annual/quarterly statements, voice response systems, and online service portals. To the policyowners, they are like the legends of the old west: repeated over and over again, year after year, so that they become indistinguishable from fact.

Unfortunately, in these cases, you cannot just “print the legend.” Once identified, insurers are required to communicate and correct these differences, which is often easier said than done. A common approach is addressing all issues that are not in the favor of the policyowner to make them whole and correcting the system to process correctly form the point of migration onward.

When the error is in favor of the policyowner, many insurers allow the customer to keep the discrepancy. The rationale is that it may be less costly than correcting the error back to inception and more palatable to customers and agents. However, this approach may not pass muster with insurance commissioners who expect all policyowners to be treated consistently. In any event, problems where the differences result in a significant financial impact to insurers present them with challenging decisions and can sometimes result in delays to the ultimate migration to the new system.

Preparing for Conversions

To head off these situations before embarking on a large conversion program, insurers should take stock of their existing blocks of policies, especially the more complex older blocks. They should review agent/customer complaints, error logs, and lists of enhancements for future processing. They should engage the subject matter experts for these blocks in knowledge-transfer activities, since many issues are undocumented and only known by long-term employees. This has become more critical as these experts may be departing sooner than expected due the increase in early retirements related to the pandemic.

In another recent approach, some vendors have developed migration services that are able to identify policies which are not operating as designed in the existing admin systems. These vendors have tools that allow them to simulate all activities from inception to provide an authoritative version of the truth for all policies.

One benefit of this approach is that it allows the conversion analysis to be performed without the new system needing to be implemented, configured, and tested. This can mitigate the risk of late discovery, which can derail the final stages of a PAS transformation and result in the need to keep the old system running in production longer than planned.

For related research on this topic, read the Executive Briefs Life Core Systems Conversion: Key Issues and Best Practices and Block Analysis and Life Systems Transformation.