Mercer Touts Use of Industry-Specific Mortality Assumptions

Society of Actuaries mortality tables are not required to be used by defined benefit plans, and there may be something better.

The recently released mortality tables from the Society of Actuaries (SOA) are not required to be used for defined benefit (DB) plan accounting and calculations, and Mercer contends there may be better assumptions for DB sponsors to use.

Jim Berberian, chief architect of the recently completed Mercer Industry Longevity Experience Study (MILES), noted in a recent webcast that auditors say the mortality assumptions a plan uses should be backed up by reliable, accurate data.

Mercer believes longevity could vary significantly among industries, said Bruce Cadenhead, chief actuary at Mercer in New York, during the webcast. Using industry-specific data rather than a generic study can reduce the risk of undervaluing or overvaluing liabilities, enhance future estimates of liabilities and cash flows, and improve reporting, he contended.

More than 200 plans were used in the Mercer Industry Longevity Experience Study (MILES), and they were divided by industry based on the business code used on Form 5500. According to Cadenhead, the study found, just as he contended multiple studies have shown, that blue collar mortality is higher than white collar mortality. He presented graphs that showed, for males, MILES found a 25% difference between the highest and lowest mortality between specific industries. For females, it found a 20% difference. For example, he explained, employees in auto/industrial goods have lower longevity (higher mortality) than employees in banking and finance.

Cadenhead said DB plan sponsors can use the SOA tables, but should determine if it is using an average experience that is unrelated to many employees. Plan sponsors can use the MILES tables or they can develop their own assumptions based on their employee demographics, but Cadenhead noted that developing assumptions is the most costly and time-consuming option, as the employer must use someone with sufficient expertise to do so.

Berberian pointed out that the SOA study showed a substantial increase in longevity, which will result in a substantial increase in pension liabilities relative to the assumptions currently in use. Plan liabilities may increase by 5% to 10% depending on the characteristics of the plan and all assumptions used, but net liabilities recognized on balance sheets could double, according to Berberian. In some cases, he said plans with a net surplus could even have a deficit using new mortality tables.

David Weissner, Midwest market business leader with Mercer in Chicago, said plan sponsors should know all the options they have for mortality assumptions and understand that small changes in liability can have a big balance sheet impact. “Assumptions must be a company’s best estimate; would a company- or industry-specific table be better than a generic table?” he queried.

He added that documentation will be a big thing in 2015—it should already be a part of plan sponsors’ processes, but new mortality tables make it especially important.

“This is your opportunity to own your assumption process and make sure generic assumptions are not forced on you,” he told plan sponsors.

The MILES report and tables may be purchased from Mercer.