The way that retirement plan contributions are allocated to employee Participants under a traditional pension or profit sharing plan is fairly straight forward – and restrictive. Such a plan would simply allocate, say, 15% of each employee’s compensation, including the doctor’s. In contrast, a "Cross Tested" ("New Comparability") plan can have different allocation rates for different classes of employees – as long as the plan otherwise pass certain tests. Such a plan might be able to allocate, say, 5% to the staff and 20% to the doctor – quite a difference! Needless to say, these plans are quite sophisticated and must be tailored to fit each office, depending on their staff statistics, contribution level, etc. This article is intended to give you a very general overview of how such a plan works.

    Benefits: The "benefit" rules are satisfied by essentially contributing the same percentage of each Participant’s compensation. Thus each employee gets the same "allocation rate". A slight discrimination in favor of highly compensated employees, typically the doctor, is permissible by integrating the plan with social security.

    Coverage: The "coverage" rules are satisfied if at least 70% of non-highly compensated employees, the staff, are included as Participants in the plan. A Standard Regional Prototype plan satisfies the 70% test because everyone who is eligible must be included, or 100%. Everyone over age 20 who works more than 1,000 hours in a year is eligible. Once an employee becomes a Participant, they can only be excluded if they work less than 500 hours in the year they terminate.

    The above tests are considered to be "safe harbor" tests that are guaranteed to be acceptable. However, like most safe harbor tests, they are fairly conservative and restrictive. "Cross Testing" is a more sophisticated acceptable alternative that can be used to push the limits of discrimination and often give the doctor/owner a proportionately higher benefit. Each plan must be individually designed – and approved by the IRS – to pass these tests based on each practice’s particulars. (Although the theory behind these plans has been around for a few years, it has only been recently that the IRS has come out with definitive regulations that have allowed pension professionals to use these plans with certainty and comfort.)

    Cross testing is based on two alternative concepts to pass the "benefits" and "coverage" tests: the use of "benefit rates" instead of "allocation rates" and dividing the plan into separate "rate groups" for testing. For the technically inclined, a brief overview follows. The rest of you may skip to Section III!

    Benefit Rates: If each employee gets, say, a contribution allocation equal to 5% of their compensation (like a traditional plan), then their "allocation rate" is simply 5%. However, a plan is allowed to convert each employee’s simple allocation rate to an equivalent or accrual "benefit rate". This is the projected future retirement value of the contribution at age 65. The plan is then able to test those future benefit rates for discrimination, instead of the simple allocation rates. If the non-highly compensated employees’ share of the total future benefits is greater than 70%, then the plan passes. Since the plan is testing the future benefit rates, always having a few employees younger than the doctor is extremely helpful, if not critical. Since younger employees have more years to project their benefit to age 65, their benefit rates can be disproportionately large in comparison to their current allocation rate. Their large benefit rates make it easier to pass the tests, while giving the doctor a much larger contribution allocation. (As opposed to an "age-weighted" plan, older employees in a cross tested plan are not harmful to the testing.)

    Rate Groups: Participants are collected into "rate groups" according to their benefit rates. Specifically, a rate group is established for each Highly Compensated Employee ("HCE"). Everyone with a benefit rate higher than the HCE’s is in his group. Each rate group then must pass the "coverage" test as though each rate group was a separate plan. A rate group typically fails the traditional, safe harbor 70% "ratio test", so it must rely on an alternative, two-pronged "Average Benefits Test". This test consists of a "nondiscriminatory classification test" and an "average benefits percentage test".

    The nondiscriminatory classification test requires that the proportion of nonhighly compensated employees ("NHCE") in each group exceeds a certain percentage published in IRS tables. (It is similar to the 70% ratio test, but with much smaller percentage requirements that change based on the employee mix.) The average benefits percentage test states that the average benefit rate of all the NHCEs must exceed 70% of the average benefit rate HCEs.

    Yearly changes in the contribution allocation formula, the total contribution and the employee mix simultaneously change all the above benefit rates, rate groups and test results. In fact, we had to develop our own computer program to handle the calculations! The above factors must be analyzed and rebalanced each year to maximize the doctor’s benefit while still passing the tests.

    As long as a plan can pass the above tests, it has the following unique possibilities that a prototype plan does NOT have:

1. An employee may elect NOT to participate in the plan. This would rarely be used for a staff person, since the employee would often expect to be compensated an equal amount as a bonus. Thus the practice does not save any money, and using this tool reduces the flexibility to use the next few benefits. However, this is a tremendous planning tool for an associate or younger owner/doctor, especially if they are a "highly compensated employee" (HCE). To be an HCE, the employee must earn over $80,000 a year OR own at least 5% of the practice. Many younger doctors would rather take home extra compensation rather than be in a retirement plan for awhile. As an added benefit, if an HCE elects out of the plan, it has a very favorable impact on the various tests above that rely on NHCE and HCE ratios.
   
2. An employee may be required to still be employed at the end of the plan year in order to get a contribution (vs. the 500-hour rule in the year of termination mentioned above).
   
3. An employee may be required to continually requalify each year by having to work 1,000 every year, rather than just the first year they initially qualified to participate. This might be used to exclude someone who drops from full time to permanent part time under 1,000 hours. (The plan will automatically override these items 2 and 3 if necessary to maintain the plan’s ability to pass the qualification tests.)
   
4. The plan may be designed to automatically exclude a category of employees, say, hygienists. This might be risky in a small plan (less than, say, five or six employees) because, invariably, one or two employees are in transition and may not be participants for awhile. In that case it may be impossible to pass the various ratio tests without all the other employees participating. This would be the same problem with an NHCE being allowed to elect out of a small plan.

    As you can imagine, there are numerous permutations and combinations that will have to be monitored each year.