Ferenczy Benefits Law Center | We are your ERISA solution
Atlanta, GA • 404.320.1100

Solution in a Flash: Retirement Plan Correction Solution Where’s Don Draper? Correcting Coverage Failures – the 11(g) Amendment

Solution in a Flash: Retirement Plan Correction Solution Where’s Don Draper? Correcting Coverage Failures – the 11(g) Amendment

Retirement Plan Correction Solution

Where’s Don Draper?

Correcting Coverage Failures – the 11(g) Amendment

Adriana M. Starr, Esq.

Roger Sterling owns the Sterling Cooper Advertising Agency (“SCAA”), the plan sponsor of the Mad Men 401(k) Plan (the “Plan”).  To provide SCAA’s account executive, Don Draper, with more creative control, SCAA established a wholly owned subsidiary, Draper, Inc. (“Draper”) in 2014.  SCAA excludes Draper employees from the Plan to prevent administrative headaches.  Historically, the Plan passed Internal Revenue Code (“Code”) Section 410(b) coverage testing, despite the exclusion of Draper employees.  Roger submits the 2020 annual census for both SCAA and Draper to Joan Harris of Holloway Harris, the third-party administrator.  Several weeks later, Joan notifies Roger that the Plan failed coverage testing.  Joan suggests that SCAA correct the failure by adopting a retroactive amendment under Treasury Regulation Section 1.401(a)(4)-11(g).  This is commonly called an “11(g) amendment.”  Roger asks Joan to walk him through the failed coverage test and the 11(g) amendment.

What is the Coverage Test?

Qualified retirement plans, with some exceptions, are subject to Section 410(b) coverage testing.[1]  Coverage testing is required to ensure that enough nonhighly compensated employees (“NHCEs”) benefit under the Plan – particularly when employees who otherwise meet the eligibility requirements are excluded from the Plan by group or category.

Who’s the Employer?

One of the key issues for coverage testing is what companies’ census information must be included in the test.  The employees of the plan sponsor must all be included.  In addition, if the employer is part of a controlled group or an affiliated service group, the employees of any other member of the group (commonly called a “related employer”) must be included in the testing.

In this case, SCAA and Draper, Inc. are part of a parent-subsidiary controlled group.  Therefore, the coverage testing must include the employees of both companies.

For more information on the effects of controlled groups on plan administration, see our previous Solution here.

How is the Coverage Test Performed?

The coverage requirements are met if the Plan passes either the Ratio Percentage Test (RPT) or the Average Benefits Test (ABT). [See Treas. Reg. §1.410(b)-2]  Coverage is deemed to be satisfied for a Plan year if either: (a) the company does not have nonexcludable NHCEs at any time during the Plan year; or (b) the Plan does not benefit any highly compensated employees (“HCEs”) during the Plan year.

Ratio Percentage Test

The RPT is satisfied by comparing two fractions: the first takes the number of benefiting NHCEs and divides it by the nonexcludable NHCEs (the “NHCE coverage fraction”).  The second fraction is calculated the same way, but for HCEs (the “HCE coverage fraction”).  The test is passed if the NHCE coverage fraction is at least 70% of the HCE coverage fraction.  So, if all HCEs are covered (that is, the HCE coverage fraction is 100%), the NHCE coverage fraction must be at least 70%. This means that 30% of the NHCEs can be excluded from the plan without violating coverage.

But, who is an “excludable” employee?  In this context, the word “excludable” refers, not to someone who is actually excluded from the plan, but to someone who need not be included in the coverage test at all.  An excludable employee is an employee who, for the entire testing period, is not eligible to participate in the plan for the year AND meets one of the following criteria: 1) does not meet the age and service requirements; 2) is a nonresident alien; or 3) is a collectively bargained employee. An employee is also an excludable employee if s/he is a participant in the Plan, but terminated prior to the last day of the Plan year, has fewer than 501 hours, and is not benefiting under the Plan. (See Treas. Reg. 1.410(b)-6).

Anyone who is not excludable is a nonexcludable employee.

So, in its purest form, we (1) take the employees of both SCAA and Draper, Inc.; (2) disregard those who are excludable; (3) categorize anyone who is left as either a HCE or NHCE; (4) determine who among those employees benefits under the plan; (5) divide the number of benefiting NHCEs by all nonexcludable NHCEs; (6) divide the number of benefiting HCEs by all nonexcludable HCEs; and (7) compare the fractions.


Who is Benefiting?

There are commonly three possible types of contributions in a 401(k) plan: the employees’ salary deferrals; the employer matching contribution; and the employer nonelective contributions.  The rules require that “benefiting” be determined differently for each of these types of contributions.  Therefore, the rules also require that coverage be separately tested for each of these types of contributions.  [See the “mandatory disaggregation” rule in Treas. Reg. §1.410(b)-7(c)(1)]

In our current case, the SCAA Plan permits deferrals and provides a discretionary matching contribution. Therefore, we will have two coverage tests: one for the deferral portion and one for the match portion of the Plan.

Employees are considered to be benefiting for the salary deferral portion of the Plan if they are eligible to defer at any point during the Plan year (regardless of whether they actually did so). [Treas. Reg. §1.410(b)-3(a)(2)]

Employees are considered to be benefiting for the match portion of the Plan if they would receive a matching contribution if they deferred during the Plan year (whether or not they actually did). [Treas. Reg. §1.410(b)-3(a)(2)]

For plans that have nonelective contributions, an employee is benefiting for such portion of the Plan if s/he actually receives either a nonelective contribution or forfeiture allocation during the Plan year. [Treas. Reg. §1.410(b)-3(a)(1)]

So, Let’s Do It!

In this case, the coverage test for both the deferral and match portions of the SCAA Plan are the same, and are as follows:

A Benefiting Employees (SCAA) 5 3
B Nonexcludable Employees (SCAA) 5 3
C Nonexcludable Employees (Draper) 8 1
D Total Nonexcludable Employees (B + C) 13 4
E Benefiting Percentage (A/D) 38.46% 75%
F Ratio Percentage (E for NHCEs Divided by E for HCEs) 51.28%

If the eligibility requirements for the match were different than for deferrals (e.g., if the plan had immediate entry for deferrals, but a 1-year wait for the match), the excludable and nonexcludable employees would be different and the tests would not be the same, as they are in this case.

Because the Draper employees are not benefiting and they represent such a large percentage of the nonexcludable group, the ratio percentage is less than 70%. Therefore, the test is failed.

The Plan may satisfy coverage by passing the average benefits test, which is more complex.

Average Benefits Test

To pass the ABT, the Plan must satisfy the nondiscriminatory classification test and the average benefit percentage must be at least 70%.  The average benefit percentage is the actual benefit percentage of the NHCEs in the testing group divided by the actual benefit percentage of the HCEs in the testing group for that testing period.  There are many different ways to calculate the benefit percentages under this test, and the flexibility is one of the reasons why it is complex.

Is the ABT Mandatory?

No.  The Plan may simply stop at the RPT and do what is needed to repair the coverage failure.  In fact, because of its complexity, the cost of the ABT may exceed the cost of correcting the failed RPT.

Joan discusses the ABT with Roger, and he opts to proceed to correct the failed RPT without further testing.

Another Trick Up Your Sleeve

If the plan has age/service eligibility provisions that are more lenient than the statutory maximums of age 21 and one year of service, there may be another way to make the coverage test pass.  In that situation, the Plan may test (using either the RPT or the ABT) those who meet the statutory eligibility requirements separately from those who meet only the lower plan requirements (these employees are commonly called “Otherwise Excludable Employees”). [See the permissive disaggregation rules of Treas. Reg. §1.410(b)-7(c)(3)]  This special rule often makes one or both coverage tests easier to pass.

As the SCAA Plan requires the statutory age and service requirements for eligibility, the Otherwise Excludable Employee option is not available.

Repairing Failed Coverage Testing

Consequences of Failed Coverage

Failure to satisfy coverage testing is a disqualifying event.  It is considered to be a “demographic” failure because the Plan has operated according to its terms, and the Plan’s document is fine … it’s just that the demographics don’t work with the plan provisions.  If there were fewer employees in Draper, Inc. to be excluded from the Plan, or if SCAA had a greater percentage of the employees, the coverage test would have passed and the Plan would be fine.  If a plan is disqualified due to the failure to satisfy coverage, Code Section 402(b)(4) penalizes HCEs by taxing the HCEs’ entire vested account.  The NHCEs’ portion of the Plan is still considered qualified and suffers no ill effect.

Bring Back Peggy – 11(g) Amendment

Treas. Reg. §1.401(a)(4)-11(g) contains the requirements to correct demographic failures – i.e., failures under Code Sections 410(b) (coverage), 401(a)(26) (minimum participation for defined benefit plans), or 401(a)(4) (nondiscrimination) – through the adoption of an 11(g) amendment.  The amendment must increase allocations or accruals for benefiting employees or grant additional employees an allocation or accrual so that enough people are benefitting to satisfy the relevant test.

In this case, we are failing coverage because too many employees are excluded.  So, what are Joan’s options for amending the plan and adding benefits for some of the excluded people?

Plan Document May Require a Certain Type of Correction: Failsafe Provisions

The Plan document may provide a “failsafe” provision that dictates how the employer should resolve certain types of demographic failures.  This type of provision dictates the method for expanding coverage to employees.  For example, the Plan may require that all nonexcludable participants share in the employer contribution, even though they were originally exempted by class.  Alternatively, it may provide that the employer only needs to include the number of people needed to meet the coverage test, but it dictates how the newly eligible participants are to be determined.  Typically, plan documents do not provide a procedure for the failure to satisfy the coverage test for the salary deferral portion of the plan.

If the plan has a failsafe provision, you need to follow it.  Failure to follow the provision is an operational failure that can disqualify the plan … so failure to do so just puts the plan in a worse position!

If there is no failsafe provision, however, the plan must be amended to fix the failure through use of an 11(g) amendment.

Joan determines that the document does not include a failsafe provision. Therefore, the Plan has significant flexibility as to what the 11(g) amendment will provide – in other words, which employees will be permitted to enter the plan so that the coverage test is passed.


The Plan must adopt the 11(g) amendment and fund the corrective contributions within 9 ½ months after the end of the Plan year (October 15th for calendar year plans).

If the coverage failure is not corrected within the permissible time under the regulation, a submission to the IRS for approval under the Voluntary Correction Program (VCP) is required if the employer wants to completely correct the failure.

Luckily, Joan kept Roger on track and the Plan is well within the time frame to correct.

Expanding Coverage

The corrective amendment must expand the group of employees benefiting to pass either the ratio percentage test or the average benefits test.  Unlike the failsafe provision discussed above, the Plan is permitted to pick and choose the employees who will receive additional allocations to satisfy coverage.  However, the amendment must have substance.  The substance requirement prohibits the Plan from granting an allocation to a “nonvested employee whose employment with the employer terminated on or before the close of the preceding year, and who therefore would not have received any economic benefit from the amendment if it had been made in the prior year.”  [Treas. Reg. §1.401(a)(4)-11(g)(4)]  The IRS clarified that a plan also may not correct a coverage failure by granting eligibility and benefits to employees with a short period of service (and, therefore, minimal compensation on which any contribution will be based). [See, IRS Memorandum dated October 22, 2004, from Carol D. Gold, Director, Employee Plans.]

In the SCAA case, the Plan can be amended to either extend coverage to some of the Draper, Inc. employees or it can extend coverage within SCAA to some employees who did not participate.  The employees at issue will be NHCEs, because adding HCEs will not help the ratio percentage (it will make it worse). The additional contributions for these employees will not be discriminatory, because they are being extended only to NHCEs.

Joan determines that the Plan must include two Draper NHCEs to pass coverage.

A Benefiting Employees (SCAA) 5 3
B Benefiting Employees (Draper) 2 3
C Total Non-Excludable Employees (taken from prior table) 13 4
D Benefiting Percentage ((A+B)/C) 53.85% 75%
  Ratio Percentage (D for NHCEs Divided by D for HCEs) 71.79%

As the ratio percentage is now greater than 70%, the test is passed.

Corrective Contribution?

Just making someone eligible for the plan is not enough. The employee must be “benefiting” so that s/he is moved to the numerator of the NHCE fraction.  Therefore, SCAA or Draper, Inc. will need to contribute on behalf of the newly eligible employees.

To correct the failed salary deferral portion of the coverage test, the company must first calculate what the employees’ deferrals would have been had they been allowed to defer for the year and elected to contribute the average rate for the NHCEs.  This is called the “Missed Deferral,” and is equal to each new participant’s plan year compensation multiplied by the Average Deferral Percentage (ADP) of the NHCEs.  While for operational failures, the employer may be able to contribute less than the full Missed Deferral, such is not the case for a demographic failure under the Treasury Regulations.  The company must contribute a qualified nonelective contribution (“QNEC”) on behalf of each of the new employees equal to that employee’s Missed Deferral.

To correct the failed match coverage test, the Plan must provide the newly eligible Draper employees with a QNEC equal to the employee’s Plan year compensation multiplied by the average actual contribution percentage (“ACP”) of the group of NHCEs who were eligible to defer for the Plan year. If the Plan uses the Otherwise Excludable Employee rules, it must apply the ADP/ACP of the group that the employee is added to.  For example, if Peggy was in the upper group (age 21 and one year of service) the Plan would use the NHCE ADP of the upper group to determine the QNEC. In this case, the ADP for the HCEs was 4% and the ACP was 2%. Joan provides Roger with the following amounts needed to correct the failed coverage test:

Employee Compensation NHCE ADP NHCE ACP 401(k) QNEC Match QNEC Total
Peggy Olson $30,000 4% 2% $1,200 $600 $1,800
Pete Campbell $45,000 4% 2% $1,800 $900 $2,700
Total $4,500

The amendment should clearly identify the employees and their respective contributions, and, if desired, may provide that the special eligibility and contribution is for the year of failure only.

Other Requirements

In addition to the requirements discussed above, the amendment must meet the following criteria:

  1. The amendment may not reduce the employee’s benefit based on the terms of the Plan in effect immediately before the amendment. For example, the Plan may not retroactively exclude an HCE from the match to pass coverage.
  1. The amendment must be effective as of the first day of the Plan year being corrected.  The effective date for the 11(g) amendment is January 1, 2020.  This means that the newly eligible employees will receive the allocation as if the amendment was in place for all of the 2020 Plan year.
  1. The amendment must not be part of a pattern of amendments to correct repeated failures. If the Plan is consistently failing coverage, amend the Plan to satisfy coverage going forward instead of adopting an 11(g) amendment every single year to correct the same failure.  This is the first year that the Plan failed coverage. Therefore, the amendment satisfies this requirement.

Roger timely signs the amendment Joan prepared and funds the contributions to the Plan. The Plan successfully corrected the coverage failure for the 2020 Plan year!


Roger knows that the Draper employees may cause the Plan to fail coverage going forward.  The 11(g) amendment is a wonderful tool to solve the coverage issue, but it should not be abused.  If there is a consistent problem with coverage, the Plan should be amended to permit some or all of the Draper employees to participate.

If you have any coverage concerns, or other questions, be sure to let us know.  We are your ERISA solution!

Additional Guidance:

Tres. Reg. 1.401(a)(4)-1

Tres. Reg. 1.410(b)-4

Tres. Reg. 1.410(b)-5


[1] Section 414(d) governmental plans, non-electing church plans, and plans that have not made employer contributions after September 2, 1974, are not subject to Section 410(b).

Print Friendly, PDF & Email
  • Posted by Ferenczy Benefits Law Center
  • On June 14, 2021