Is a graduated matching contribution plan as to tenure permissible?

[Guest Scott Mitchell]

Does anyone know if it is permissible for an employee to have a graduated matching contribution plan according to tenure with the company? For instance, paying 25% up to 4% for 2-5 years of service, 50% up to 4% for 5-10 years. etc. How will the nondiscrimination rules affect such a plan?

[Bill Berke]

Theoretically it is possible. But the reg's are quite clear that the effect of such a formula must not have the effect of discriminating in favor of the HCEs. The testing will be an annual test based upon the actual allocations in accordance with plan's formula. I think you could test under the 401(a)(4) rules. Other people may have a different answer. I also suspect that, with time, the formula will become discriminatory unless you are working with a fair sized, stable employee population.

[Jon Chambers]

It's definitely possible, used in practice, but relatively infrequent. Probably less than 1% of plans use a service graded match. I believe testing is just 401(m), you probably need to do a coverage test (410(B)) on each match rate. Otherwise, it's not a big deal.

[Kirk Maldonado]

Jon:

Your answer seems ambiguous. Do you believe that 401(m) is the exclusive testing methodology, or must you also test each level match as a benefit, right or feature? (Without doing any research, my instinct is that you have to do BRF testing.)

[Guest]

kirk is correct. there is a benefits/rights/features issue involved.

net effect is the following: lets say you have

25% for 1-4 years

50% for 5-9 years

100% for 10 or more years.

In addition to the ACP test, you now have a coverage test involving 3 plans. For example, only those with 10 or more years are teated as 'includable and benefiting' while everyone else is treated as 'includable and not benefiting' in the third group. I have one plan that ends up around 25% every year, which, fortunately is a large enough plan, that the NHCE concentration percentage (around 95%) produces a safe harbor percentage of arounf 25%. Just enough to pass.

[Guest RJT]

Technically, it is not a coverage test. It IS, as Kirk mentions, a BRF test (as specifially noted in 1.401(a)(4)-4(e)(3)(iii)(G)), which is a nondiscrimination-not coverage-test. To run the current availability portion of the BRF nondiscrim test, you use the average benefit test for each class (inclusive of those matching up to the class amount you are testing), without regard to average benefits, using 1.410(B)-5 per 1.401(a)(4)-4(B)(1). Effective availability is then a facts and circumstances determination.

[Jon Chambers]

Kirk, as you know, I'm an investment consultant, not an attorney. My point was that I do see service graded matches in the market, and that my (limited) understanding of the compliance requirements were that you had to do 401(m)/ACP testing, plus additional testing. As Tom and RJT point out, I incorrectly referred to the additional testing as "coverage", when I should have referred to it as "BRF", and indicated that it was "not a big deal". I'm sure the testing isn't easy, didn't mean to downplay it. But as long as you have the right participant population and can pass the test, the service graded match approach is a reasonable and valid plan design that may address many companies' benefits objectives. Perhaps due to the testing complexity, we see service graded matches less frequently than might be appropriate if benefit design were the only relevant consideration.

[Kirk Maldonado]

Jon:

Whether you are an investment consultanta or an attorney does not really matter on BenefitsLink. What counts in the quality of the responses.

I directed the question to you because your comments on BenefitsLink are typically ver helpful and incisive.

[AndyH]

Scott:

In my experience, these are more common with non-profits, who tend to have less HCEs and less turnover. Plus, the HCE's often have less service than the average employee.

I agree that the benefits, rights, and features testing is required. It is often easily passed with a non-profit or any company meeting the above description, but in a regular company when the owner has the longest service, it's a flunker.

[Guest FREE401k]

For what it's worth, we have a large Plan (approx 6,000 participants) with a company match based on years of service. They are not a non-profit. This particular company is in a high-turnover industry and structured their match like this to reward the employee who stays with the company. I don't get involved in testing much, but I know this Plan always passes the test (probably because it is so large, the high number of non-HCEs easily offset the relatively few HCEs).

[AndyH]

Very valid point. I should have specified that I was referring to relatively small, closely held companies.

[Bill Berke]

To FREE 401k: Which tests is the plan using to pass? I can envision passing ADP/ACP but that is not the issue. What about 401(a)(4)? Thanks