401(k) Plan Mega Roth Backdoor After Tax Contributions

[VIkram Aurora, QPA, QKC]

Hi, I've worked on 401k plans for a while now, but encountered mega roth backdoor quite recently. I do understand that it has great tax saving for an individual but after digging deeper I read that ACP must be passed for the after tax voluntary contributions even if the plan is safe harbor and passed ADP and ACP for matching. Am I understanding this correctly? Also is there any way to bypass the testing for mega roth backdoor after tax contributions?

Thanks!

[John Feldt ERPA CPC QPA]

Am I understanding this correctly? Yes.

Is there any way to bypass the testing for mega roth backdoor after tax contributions? Have zero eligible nonhighly compensated employees, no testing needed. In other words, employing only highly compensated employees who meet the age/service/entry requirements = no testing needed. 

[CuseFan]

Exactly, this is either an owner-only play or may work in a very large plan that otherwise easily passes ACP testing, but is essentially a nonstarter in small plans that cover NHCEs.

[VIkram Aurora, QPA, QKC]

Thank you!

[austin3515]

I have had it where it was an NHCE who wanted to do it.  It was a married couple, and the NHCE was married to someone who made a gazillion dollars so they were looking to contribute the full 415 limit.  And the plan sponsor was willing to accommodate (small employer).

[JonC]

We work primarily with larger plans that are unlikely to be designed as safe harbor--but that in general, don't need safe harbor protection, because a generous match encourages participation and the ACP test is passed easily. We typically see approximately 10% of a large plan population using the mega backdoor Roth feature--perhaps a couple of hundred individuals in a 2,000 participant plan (typical for our clients). That's enough to slightly degrade the ACP test results (because most using mega backdoor Roth are HCEs) but not enough to cause the ACP to fail.

The bigger issue that we see occasionally is 415 limit violations, if someone miscalculates benefits or compensation and the aggregate limit for deferrals, match and after tax exceeds the DC 415 limit. We see a handful of 415 limit violations annually on our larger plans with this feature--I'd approximate the number as about 5 per 10,000 eligible, or 1,000 using the feature. So 415 limit violations are not common, but they do happen.

[Peter Gulia]

If a plan provides that only nonhighly-compensated employees may make an employee (after-tax) contribution, does that meet coverage and nondiscrimination conditions?

(Of the employer I’m thinking about, many nonhighly-compensated employees have modified adjusted gross income, or even one’s compensation alone, that precludes the individual from Roth IRA contributions.)

Am I right that qualified-plan conditions don’t restrain discrimination against highly-compensated employees?

If my guess is right, is there another reason (beyond § 401(a)(17)’s constraint and § 415(c) limits) why a plan’s sponsor might prefer not to allow employee contributions?

[austin3515]

2 minutes ago, Peter Gulia said:

If a plan provides that only nonhighly-compensated employees may make an employee (after-tax) contribution, does that meet coverage and nondiscrimination conditions?

Yes absolutely.   You can do anything for just NHCE's pretty much, such as profit sharing of varying rates.

[Peter Gulia]

Is there another reason why a plan’s sponsor might prefer not to allow nonhighly-compensated employees to make an employee (after-tax) contribution?

[mjbais1489]

40 minutes ago, Peter Gulia said:

Is there another reason why a plan’s sponsor might prefer not to allow nonhighly-compensated employees to make an employee (after-tax) contribution?

Im on the advisor side. The downsides I see aren't testing related they are process & communication related. 

Process - this likely wont happen very often, is the payroll team strong enough for the employer to make sure its handled correctly every time it does happen (they will have to be re-explained the rules every time it happens and I wouldn't be surprised if a payroll person told an employee it wasn't possible on accident. )

Communication - You need to let employees know its possible.  Any complicating things like this muddies the water more than adds value to most employee conversations IMO.  I'm trying to get participation & savings rates up, most times adding another layer of decisions brings confusion and decreases engagement. 

 

If a client came to me and said we have an NHCE who wants this, I would add it to make the client & employee happy. I would mention in employee meetings but very quickly and move on so I avoid creating a confusing discussion. I would be on the payroll calls initially and I would tell everyone to call me if any other employee wants to do this. The participant election of after-tax on the recordkeeper site would have to be managed well too.

[John Feldt ERPA CPC QPA]

Peter, you asked: Is there another reason why a plan’s sponsor might prefer not to allow nonhighly-compensated employees to make an employee (after-tax) contribution?
 

Yes, there is. When you run an ACP test, you only include the employees eligible for the match. So, if the match allocation has conditions, like last day or 1000 hours, then those employees aren’t in the ACP test. Yay. BUT, as soon as you allow after-tax, everyone eligible to contribute after-tax is now in your ACP test, even those that did not meet the conditions for the match. That could add a whole lot of NHCEs with zeros to the ACP test.

[austin3515]

I will also add that allowing after-tax and Roth could be making it easy for employees to do something stupid, like contribute after-tax contributions without first maxing out their Roth 401(k).  That's probably the most obvious issue with adding after-tax contributions in general. 

[Peter Gulia]

mjbais1489, thank you for your helpful observations. I see how they could be worries with many employers.

For the particular employer I’m thinking about, the only worker who submits information to the payroll service provider or to the retirement plan’s recordkeeper is the same human-resources worker who manages the retirement plan’s design and administration. And she is adept at catching both service providers’ errors. I’d explain that the time of the human-resources worker in correcting the service providers’ errors might be a meaningful burden.

For the particular workforce, uses of an after-tax contribution, even restricted to nonhighly-compensated employees, would not be few.

For this plan, eligibility begins when the employment begins. Participation for elective deferrals hovers between 99 and 100%, even for new hires. There is no meeting to explain the retirement plan. The plan has no adviser. (I advise the plan sponsor on nonfiduciary points, and advise the sponsor/administrator by writing the SPD.)

austin3515, your point is something I’d usually advise, but is not a worry for the workforce that caused me to think about whether it could be feasible to allow employee contributions.

But John Feldt’s note ends my thinking about reconsidering a plan-design choice made many years ago. Even one unnecessary weak number in the ACP test could attract terrible trouble.

[AlbanyConsultant]

I have a large law firm client that recently asked about mega backdoor Roth and voluntary after-tax contributions in general.  I explained the ACP test issue and after a slight pause, one of the partners said, "but that can be resolved by making a company contribution instead of refunds, right?"  They are actually considering corrective QMACs to let their highly paid (and, coincidentally, HPI) participants do this.  Sure, it works mathematically if they don't mind spending the money, but I have to wonder if it will really work out better than adjusting their class-based profit sharing.

[austin3515]

Maybe top-paid group would help?

[Peter Gulia]

What amount is the law firm’s line between the lower 80% and the top-paid 20%.

Might the lower 80% include many or some with seven-figure compensation?

Might the lower 80% include many more with compensation between $360,000 and $1 million?

What is the firm’s mix of capital-interest partners, income partners, counsel, senior associates, beginner associates, and assistants?

[Bri]

Targeted bottom-up QNECs to the ACP test with those rules.  Sounds fun enough.

[AlbanyConsultant]

2 hours ago, Peter Gulia said:

What amount is the law firm’s line between the lower 80% and the top-paid 20%.

Might the lower 80% include many or some with seven-figure compensation?

Might the lower 80% include many more with compensation between $360,000 and $1 million?

What is the firm’s mix of capital-interest partners, income partners, counsel, senior associates, beginner associates, and assistants?

I'm definitely looking into the top paid group.  I think the line will be north of $200K, but less than $500K.  2025 is literally the first plan year, so I don't have full data yet.

What's the goal of your last question?

[Peter Gulia]

A firm’s mix of capital-interest partners, income partners, retired partners, counsel, senior associates, beginner associates, and assistants might affect:

which workers seem likely to desire an opportunity to make employee (after-tax) contributions;

how student debt affects a worker’s capacity to make contributions;

how the expenses of QMACs are spread—for example: to capital-interest partners only? or to an income partner to the extent, wholly or partly, an associate or assistant in her income-measured practice gets a QMAC?;

how much or how little leverage affects partners’ capital and profits interests;

how the firm’s obligations to retired partners affects the firm’s financial capabilities;

how a retirement plan’s design and features affects workers’ perceptions about the firm.

[AlbanyConsultant]

I see.  Those are good points.  Thanks!