Evaluating retirement readiness formulas

Retirement plan sponsors should evaluate the assumptions used by providers in their retirement readiness calculator formulas. This can result in more accurate projections that help participants make better long-term savings decisions. A recent PlanSponsor article quoted Milliman consultant Kevin Skow discussing some assumptions that sponsors need to assess to improve retirement readiness projections.

Here is an excerpt from the article:

To calculate these projections, providers have to make numerous assumptions about key variables. Some, such as future inflation rates, do not relate to individuals specifically. But many variables do, and the default assumptions a provider uses may or may not reflect reality for a plan’s participants. “In our mind, the assumptions made are critical,” says Kevin Skow, principal and consultant at Milliman Inc. in Minneapolis.

In order to evaluate readiness formulas, plan sponsors should start by looking at three key areas where assumptions are made:

Salary, retirement date and savings. Understand what salary-increase rate a provider’s model assumes, Skow recommends. “How does that equate to what’s happened historically at your company, or what is anticipated in the future?” he says. And these models assume an average retirement age in doing the calculations, he says, so in evaluating a provider’s model, it helps to know whether that number reasonably lines up with a work force’s actual retirement patterns.

The models also hypothesize about a work force’s retirement-savings rates, going forward. In its retirement readiness calculator for participants, Milliman actually asks each to input any deferral increase he plans. When it does plan-level reports on retirement readiness, the company typically takes a “snapshot” approach and does not assume a deferral-rate increase by participants, Skow says. “But if a plan has automatic increases, a model could assume that everybody who was auto-enrolled at a 5% deferral with a 1% increase, for example, will stay with it,” he says. “Most people who are auto-enrolled stay, and very few tend to opt out….”

Additionally, retirement readiness models have to make assumptions about how long people will live… The suppositions about how long people will live have a big influence on these calculations, Skow says. “In our tools, we tend to project that an individual will need income until age 95 if that person is male, or 97 if that person is female,” he says. “Many models use the normal mortality rate in the U.S. today, which is in the late 80s.” A model assuming a shorter lifespan will improve someone’s monthly retirement-income projection, but also may create false security for some who then end up outliving their savings.

Leave a Reply