The challenge

A long-standing Milliman client that maintained a complex pension plan was looking to modify their pension plan formula. They wanted to maintain the current benefit structure and to provide new participants with benefits under a different pension formula. They also wanted to introduce cost sharing into the plan in the form of employee contributions. This was further complicated by the client’s decision to amend the plan mid-year-- specifically, four months into the plan year.

Under the current structure, participants were in one of two tiers based on their date of hire. Tier I participants had accrual rates that varied with service based on compensation for the prior 12 consecutive months; how those rates varied depended on a participant’s age and service at retirement. Further, an alternative minimum benefit was applied so that the aforementioned formula would satisfy the requirements of Section 411(b). Tier II had accrual rates that varied with service based on the average compensation for the highest prior 36 consecutive months.

Under the amended plan, benefits earned prior to the amendment would continue to grow based on increases in compensation, though no further service would be applied. The benefits earned after the amendment would have different optional form factors and would include employee contributions. For Tier I, the benefits earned after the amendment would now be based on average compensation for the prior 36 consecutive months instead of 12 months, have an early retirement factor applied, and have different retirement eligibility criteria.

The challenge was to demonstrate that the plan not only currently passes nondiscrimination testing but also would pass in the future. Therefore, it was decided that testing should include the impact of the amendment.

The solution

First, in order to have the post-amendment benefit more fully influence the testing results, it was decided to change from the accrued-to-date method to the annual method. In this way, the benefit accrual being tested was one-third based on the pre-amendment formulas and two-thirds based on the post-amendment formulas.

Second, two valuations were performed at the beginning of the plan year. One used the existing benefit formulas without regard to the amendment and the other used the updated benefit formulas reflecting the amendment. In both bases, the accrual for the year was calculated.

The year’s accrual reflecting the amendment, less an annuitized form of the employee contributions for the year after the amendment, was used to determine the Normal Accrual Rate.

One-third of the year’s accrual using the existing benefit formulas without regard to the amendment was considered to be the pre-amendment accrual. The year’s accrual reflecting the amendment, less the pre-amendment accrual, was considered to be the post-amendment accrual. For each of these pieces, the earliest age at which the benefit could be taken was determined, the appropriate early retirement and joint and survivor factors were calculated, and an interest adjustment to the retirement age was calculated. From these, the Most Valuable Factor was determined and applied to each piece. The resulting total, less an annuitized form of the employee contributions for the year after the amendment, was used to determine the Most Valuable Accrual Rate.

Unfortunately, after all of that, the plan was not passing general testing under 401(a)(4) because the minimum ratio percentage still fell below the 70% passing threshold even after applying permitted disparity. However, the client also has a defined contribution (DC) plan and all of their employees are both participants in and benefit from both the defined benefit (DB) plan and the DC plan. As such, we were able to use the Average Benefit Percentage Test to test the DB plan on a standalone basis as provided under 1.410(b)-5(e)(3)(ii). Having passed that test, we were able to compare our minimum ratio percentage of 46.14% to the midpoint of the Safe Harbor Percentage and the Unsafe Harbor Percentage, which was 45%, and pass general testing.

The outcome

The client’s plan passed nondiscrimination testing, which pleased everyone. Further, it did so in a way that incorporated the plan’s complex mid-year amendment. This gave the plan sponsor greater confidence that the plan will continue to pass nondiscrimination testing after the transition year.