How I would redesign a shared DB/DC retirement program

Executive summary:

  • We see both advantages and disadvantages to IBM's retirement program changes. One advantage is that more capital will be freed up for other corporate initiatives, while one disadvantage is that without a 401(k) match, participants may save less for retirement.
  • We believe a more optimal shared DB/DC retirement plan would reduce, but not eliminate, the 401(k) matching contribution.
  • We also think it would be better to tie the interest on the notional account balance to actual portfolio returns or a common market index rather than a Treasury yield like IBM has done. We would also offer participants the option of a bifurcated benefit in addition to the life annuity option.

IBM made headlines several weeks ago when they announced groundbreaking changes to their employee retirement program, essentially reopening their long-frozen defined benefit plan to replace their 401(k) plan matching contributions. Industry experts have hypothesized on the motivations behind this change, ranging from the cash advantages to the shared risk model of combining tax-advantaged defined benefit (DB) and defined contribution (DC) plans.

So far, it is unclear if this change will spark a new wave of reopened DB plans or if it will be a quick flash in the pan. If nothing else, IBM's change sparked renewed debate on the value of pension surplus. Plan sponsors of overfunded plans, or those just developing their endgame DB plan strategies, ought to take notice and consider the potential advantages of reopening their DB plans. While few sponsors are in the same situation as IBM, lessons can be learned from their strategy.

As a quick refresher, IBM made the following retirement program changes:

  • Eliminated the employer 401(k) plan matching contribution (plus a small nonelective contribution).
  • Introduced a new notional account balance benefit in the DB plan (previously frozen since 2008) that offers 5% percent of pay each year plus interest. The benefit is payable at the time of termination or retirement. This benefit is "notional" since it only exists on paper until it is paid (just like any other DB benefit, by the way).
  • Interest on the notional account balance will be 6% through 2026, then be based on the 10-year Treasury yield, floored at 3% until 2033.
  • This new benefit will be funded through existing U.S. DB plan assets, which hold a $5 billion surplus above current U.S. qualified DB liabilities. IBM will likely not be required to put actual cash in the DB plan for many years to come.

While not everyone will agree with how IBM designed this change, I see some advantages:

  • For IBM, this is an undoubtedly advantageous use of accumulated pension surplus that is likely to save over $500 million per year in cash contributions in the near term, freeing up capital for other corporate initiatives.
  • For the participant, this reduces investment risk and offers an annuity option within the DB plan that could reduce longevity risk. Portability of the benefit to an IRA or other 401(k) is still available.

That said, I also see some disadvantages:

  • For IBM, as many as 60,000 participants will be introduced into the DB plan, increasing PBGC flat-rate premiums (over $100 per participant in 2024) and other administrative costs. However, the majority of these costs can be covered with the existing plan assets.
  • Participants lose the ability to generate investment returns on the employer contribution, which could be far better than the 10-year Treasury yields offered after 2026. They also lose the incentive to contribute to their 401(k) plan by taking away the employer match, which I expect will negatively impact the participants' likelihood of sufficiently saving for retirement.

Since this announcement was made, I have been considering alternative designs for a shared DB/DC retirement program using a previously closed and frozen DB plan. While I believe the reopening of the DB plan was a step in the right direction to address meaningful retirement challenges, I think some adjustments could help crystallize the benefits of this type of arrangement.

Three enhancements I would make to IBM's newly-announced retirement program

First of all, I would have reduced, not eliminated, the DC matching contribution. When sponsors eliminate the DC match, employees lose a significant incentive to contribute themselves. Additional employer contributions also provide more assets for the employees to work with and invest in. You could reduce the employer match to 50% of the first 5%, down from 100% of the first 5%. This could be paired with a reduced pay credit within the cash balance benefit of about 3% to keep the benefit similar to what it was before the plan change. The cost savings and use of pension surplus would still be significant (and potentially last longer) while maintaining the incentive for participants to keep saving.

Second, in the DB plan, I would tie the interest on the notional account balance to actual portfolio returns or a common market index rather than a Treasury yield. This would likely provide better returns to the employee over time, but it also helps participant and employer share in the investment risk and allay concerns about earning less than the DC plan would have. Safeguards such as floors and caps help smooth out the return pattern and protect the participant against large losses right before (or during) retirement. These types of interest crediting rates within cash balance plans have become increasingly popular among sponsors of small offices and law practices but have yet to take wide hold with corporate sponsors.

Third, in addition to the life annuity that must be offered within any qualified DB plan, I would provide the option for a bifurcated benefit, i.e., the ability to take a portion of the benefit as a lump sum (to be transferred to an IRA or 401(k) plan). At the same time, the rest is received in annuity form. This improves plan participant options at retirement depending on their circumstances. For example, they could take 50% of the benefit in the form of a lump sum to boost their retirement savings, while receiving 50% as a cost-effective annuity option to help alleviate longevity concerns.

The bottom line

I believe these plan design changes would strike a reasonable balance between employer and employee risk and responsibility, retain employee contribution incentives, and take advantage of markets to generate returns. Improving funded status across the DB industry creates exciting opportunities to explore uses of pension surplus beyond plan hibernation and plan termination. The discussion is just getting started.

Download our research paper