Can the document be silent as to the correction method and can differe

[John A]

In cross-tested Plans, should the document include or exclude the method to correct the test? If the document does not specify the correction method, can different correction methods be used for different years?

[Lorraine Dorsa]

Most cross-tested (and most non cross-tested, for that matter) plans do not specify the correction method to be used if any of the non-discrimination rules are satisfied, and there is no requirement that they do so.

Our firm does a lot of these plans and it is our strong preference that the document NOT specify the testing or corrective methods.

If the document does not mandate methods, we can use all possible options for testing and, if we do fail, choose the correction method which is most cost-effective and best meets the employer's needs. Each year is independent--you can select the testing and correction method which works best, regardless of the method you selected last year.

We admininister some plans which specific testing and/or correction methods and they can extremely frustrating--sometimes we can pass the tests using other than the method specified but cannot since we must follow the terms of the document; other times we can make a simple correction to pass but cannot since again we must follow the terms of the document.

------------------

[AndyH]

I'm very surprised to read the response that the correction method need not be in the document. Clearly the testing methodology does not need to be in the document, but I wonder how excessive "employer discretion" is avoided in the correction process. Is there no requirement that allocations be definitely determinable?

I'm not saying I disagree, just that I find this very surprising and contrary to what I thought I'd heard at conferences, and contrary to the position our legal department takes (which may or may not be needlessly conservative.)

I thought a corrective amendment would be needed by the following 10/15.

Any elaboration on how a correction method can be changed from year to year without it being documented through a plan provision or amendment?

[mwyatt]

Maybe I'm missing something, but now that the IRS is allowing a CT plan document to define the allocation method as allocation by defined groups with the employer providing written instructions to the Trustee of how to allocate the total contribution by group, doesn't this problem go away? Obviously the written instructions that you as the consultant provide to the Employer satisfies the General Test (or they better ).

I thought that this was the most powerful change by the IRS as you no longer had to worry about your "super integrated" or specified allocation method failing from year to year, with corrective (nondeductible) contributions being made to pass the General Test.

[Lorraine Dorsa]

To Mywatt: Yes, in plans in which the employer specifies the amounts that are allocated to each class the testing should be done before the decision is made re the amounts to be allocated and therefore testing should not be an issue. However, not all plans are class plans; we still have a lot of super-integrated and other types of formulas out there.

To AndyH: The correction is not discretionary in the sense that the employer randomly allocates $ to the people they like such that the tests pass. A corrective amendment which satisfies 1.401(a)(4)-11(g) is required. However, there is a lot of flexibility under -11(g) and therefore a correction method which best suits the needs of the employer, within the constraints of -11(g), can be selected. And since the corrective amendment applies only for a given year, the corrective amendment, if any, for next year can be different.

------------------

[AndyH]

To Lorraine: Agreed. Thank you for the clarification.

To John A: Clearly, different firms have different approaches. We try to insert "failsafe" language which is intended to be permanent, which increases individual dollar allocations if necessary (starting with the youngest) sufficient to elevate NHCE's enough to pass the a(4) test. This avoids many headaches, and the correction is usually inexpensive.

The IRS is inconsistent in it's acceptance or rejection of this, but we usually get it through.

If this correction is needed more than one year in a row, we consider structural redesign if it does not project to pass in the future.

[This message has been edited by AndyH (edited 04-13-2000).]