Freezing a defined benefit (DB) pension plan has become common practice over the past decade. Plan sponsors give many reasons for freezing the DB plan, but one of the most common is that the funding requirements are too expensive and volatile. In a recent article, two actuaries from Milliman dissected a sample pension freeze and argue that this tactic is not always cheaper for the plan sponsor, and often provides a worse benefit for participants.
This post does a quick analysis of the article, but I’d encourage you to read the entire document (5 pages) for all of the details. The basic set-up of the case study is this:
- A DB plan with fairly rich benefits is underfunded and subject to large IRS minimum required contributions. So, the plan sponsor decides to freeze benefit accruals in the DB plan in order to limit their exposure to future liability accruals.
- In order to soften the effects of participants of the DB freeze, the 401(k) defined contribution (DC) plan benefit is increased from a 50% match on 6%-of-pay (i.e., maximum 3% company match) to a maximum 401(k) match of 4%-of-pay along with an additional non-matching contribution of 4%-of-pay (i.e., increase from maximum 3%-of-pay to maximum 8%-of-pay DC benefit).
- Maintain current DB plan asset allocation
The article goes on to propose a different course of action:
1. Continue accruals in the DB plan, but have future accruals be based on a less-expensive benefit formula (e.g., get rid of early retirement subsidies and move to a career average pay plan).
2. Immunize future benefit accruals in the DB plan using a Liability-Driven Investment (LDI) strategy.
One of the authors’ main arguments is that the problem with the DB pension plan is not necessarily future accruals, but rather it is the cost of paying for the unfunded past-service benefits. Funding these accrued benefits cannot be avoided by switching to a DC plan. As long as the plan sponsor is stuck with financing this shortfall, why not provide an “optimal” benefit for future accruals (i.e., a DB benefit since it provides lifetime income and less investment risk for participants)?
I agree that it’s possible to reduce the DB plan’s interest rate risk on future accruals by implementing LDI, but here are some other considerations:
- If the pension plan is severely underfunded, is it wise to continue DB benefit accruals (even modified ones) when you can’t even afford the benefits that have been accrued to date? Moreover, money spent to enhance a DC plan would probably be better spent shoring up the DB plan rather than providing new benefits. A DB pension freeze may be prudent in this case, and it allows the plan sponsor some time to carefully consider whether the DB plan will continue to meet its objectives.
- Does the plan sponsor really want the administrative complexities and costs of a DB plan? These won’t be eliminated by a pension freeze, but having a two-tier pension benefit can be confusing to participants.
- Is the DB benefit an effective means of attracting and retaining talented employees? For young employees I’m not convinced that this is true in many cases. However, the recession has highlighted some of the benefits of DB plans (at least for employees) and there may be a new appreciation for them among mid- and long-service employees.
- Is the scenario presented in the case study really applicable to many plan sponsors? Enhancing a DC plan is a common reaction to freezing a DB plan. However, the replacement DC plan in the case study is very rich compared to what many plan sponsors might consider offering their employees in these lean economic times. The article’s argument that cost (and volatility) of an adjusted DB plan is comparable to a frozen DB plan with (greatly) enhanced DC benefits may rely on an aggressive interpretation of what type of DC enhancements will really be offered.
Overall, I think this article does a nice job of highlighting some important details that are often overlooked when plan sponsors decide to freeze their DB plan. I’ve heard many stories of sponsors who are surprised when their pension costs don’t go away after a freeze, so they need to be made aware that the real cost reduction is only for future benefit accruals.
I also like the authors’ suggestion that a viable alternative for many plans is just to reduce the richness of future DB pension accruals instead of turning them off altogether. It might be a bit of a pain to keep track of a “frozen benefit” as well as a “new accrual benefit”, but it might not be too bad.
Lastly, the article describes a key benefit of DB plans to retirees: the “85-year-old test”. This encourages us to look at DC and DB benefits from the perspective of an older retiree. If we were 85 years old and relying on employer-provided benefits to support our standard of living, then would we rather have the assurance of a life-long Defined Benefit annuity, or the unknown quantity of a 401(k) balance which may or may not be exhausted by the time we are 85 years old?