Comments With Respect to Prop. Treas. Reg. §§ 1.401(A)(4)-8(B)(1) and 9(B)(2)(V)
Governing So-Called “New Comparability Plans”
March 23, 2001

Contents | Introduction | Summary | I | II | III | IV | V

Analysis

III. “Broadly Available Allocation Rates” Requirement

Under the broadly available allocation rates requirement, allocation rates for age or service weighted plans must increase smoothly and at regular intervals. A schedule of allocation rates has regular intervals of age or service if each age or service band, other than the band associated with the highest age or years of service, and any band beginning before age 25, is the same length. This test further requires that there be no more than a 5% increase between the bands, the ratio of rates between bands not exceed two and the ratio of rates between bands not exceed the ratio between the two immediately preceding age or service bands.

We suggest that the mathematical exactness of the proposed "smoothly increasing" and "regular intervals" in the broadly available allocation rates requirement is overly restrictive, could add great complexity to plan administration, and could in practice result in lower contributions for many participants at younger ages.

Effect on Younger Employees. The proposed complex set of constraints will allow plans to have significant contribution differences if rates are changed often enough and begin at low contribution levels. In the following examples, a plan can pass these rules and allow the oldest employees to get a contribution more than 15 times that of the youngest employees. But these requirements could cause problems for plans that are less discriminatory in favor of HCEs, even though younger employees receive a more meaningful level of contributions (with fewer contribution rate bands as they age).

The schedule in Plan B (an existing plan) does not meet the broadly available allocation rate test even though it provides a more meaningful benefit to younger (and presumably lower paid) employees than the schedule in Plan A (a hypothetical plan), which would meet the test.²

The first problem under the test for Plan B is that the increase in contribution rates at age 30 is 50% (as the contribution rate moves from 10% to 15%), while the increase at age 25 is only 25% (8% to 10%). In order to conform to the requirement that the percentage increase in contribution rates cannot go up at higher ages, the plan could reduce the contribution at ages under 25 to 5% of pay. Clearly this is an undesirable solution that will harm younger employees, who also tend to be NHCEs.

The second problem with the schedule in Plan B arises from the irregular intervals (25–30 vs. 30–45). This could be solved by reducing the contribution rate to the age 30–35 group from 15% to 10%, making the 25–35 interval an even 10 years. Again, those in the age 30–35 group, many of whom will be NHCEs, will receive less than they would under the plan before the change.

Plans with Multiple Formulas. Another way in which the broadly available allocation rates requirement seems overly restrictive is that it makes no accommodation for grandfathered benefits or for different allocation rates for different profit centers or geographic locations. In our experience, it is not uncommon for defined contribution plans to have different allocation formulas for grandfathered groups as a result of plan mergers following acquisitions or similar transactions, or to have the same allocation formula for all participants but base total contributions for groups at different locations or profit centers on that location’s or center’s profitability (resulting in different allocation rates for each group). Such plans typically pass the general test easily using cross-testing, but in many cases would be unable to satisfy the broadly available allocation rates requirement and therefore could effectively be prohibited from using cross-testing under the proposed regulations.

We believe that, to address this problem, at a minimum the broadly available allocation rates requirement should be modified to (a) disregard limitations on the availability of allocation rates that result solely from grandfathering restrictions imposed in connection with acquisitions or similar transactions, and (b) disregard differences in actual allocation rates that result solely from differences in contribution levels for groups at different locations or profit centers, e.g., based on the profitability of the location or center.

If the IRS and Treasury decide to keep this structure, we recommend that the broadly available allocation rates scheme provide for more flexibility to recognize the diverse types of designs that currently exist. The focus should be on the upper ranges of age and service and only if a plan design blatantly discriminates in favor of highly paid individuals should it fail to meet the requirement. In our experience, there are no plans that are currently structured as the proposed regulations would require.

Contents | Introduction | Summary | I | II | III | IV | V