Are RMDs triggered once a distribution is taken?

[asdfgdfa]

Participant is older than 70.5 and still working. The plan only allows withdrawals with an RMD, so if they take a withdrawal this year they will also take an RMD. If they take a withdrawal + RMD this year, and continue working next year but don't take any withdrawals next year, will an RMD be required next year? are the RMDs permanently triggered? thanks.

[duckthing]

Maybe I'm missing something in your explanation... but assuming this participant wasn't a >5% owner the year they turned 70.5, it sounds like they want to take an in-service withdrawal that's not permitted by the plan. Taking one anyway and calling it an RMD doesn't make it an RMD.

[asdfgdfa]

38 minutes ago, duckthing said:

Maybe I'm missing something in your explanation... but assuming this participant wasn't a >5% owner the year they turned 70.5, it sounds like they want to take an in-service withdrawal that's not permitted by the plan. Taking one anyway and calling it an RMD doesn't make it an RMD.

The plan allows for in-service distributions in excess of the RMD.

[duckthing]

Understood. Is this participant required to take an RMD? Or are they just trying to take an in-service withdrawal and thinking that if they call part of it an RMD they'll be able to take the rest as well?

[ESOP Guy]

I am with duckthing and/or your description is unclear.

What exactly does the plan say?  Does it say that someone who has to take an RMD can take a distribution in excess of the RMD also?   If so, than doesn't the person have to be required to take an RMD before they can take a withdrawal?   You seem to be saying a withdrawal can trigger an RMD.  I am not sure how that works as that isn't a requirement under the law for an RMD. 

 

[Luke Bailey]

Would really need to see the plan language, but it sounds like the RMD for year is $0, so if they take the in-service, maybe they are taking the RMD.

[duckthing]

3 hours ago, Luke Bailey said:

Would really need to see the plan language, but it sounds like the RMD for year is $0, so if they take the in-service, maybe they are taking the RMD.

I could be misunderstanding the question here. It sounds to me like the participant has not yet reached their required beginning date but wants to take an in-service withdrawal that would not otherwise be permitted by declaring part of it to be an RMD.

It sounds to me like the original question boils down to "If this participant chooses to take an RMD this year, can they choose not to take one the following year?" And I don't think the premise of the question makes sense -- either the participant is required to take an RMD, or they're not.

[Doc Ument]

 

I agree with everyone who posits that RMDs are never optional, which is why the R stands for Required.

 

The question is: Is the distribution being characterized as “required” by reference to the plan’s language enforcing the Code or is it “required” by some plan provision other than the plan’s language enforcing the Code?

 

If I understand the original question correctly (the identification of which seems to have become the primary topic), the law does not require RMDs simply because a participant chooses to withdraw funds from a plan after attaining age 70½, though the plan might do so just because its designer thought it was a great idea. The law’s only “trigger” is either age alone (i.e. age 70½ for everyone if the plan still uses the pre-SBJPA provision), or (for a plan that uses the post-SBJPA provision) a combination of age, ownership status, and employment status. Only distributions mandated by the Code are “true” RMDs. For example, such amounts are automatically not eligible for rollover if they are required by the Code. In contrast, if the distributions are required only because the plan is fabricating “pretend” RMDs as described below, then such distributions would not be required by the Code and such amounts could not (for example) therefore be automatically characterized as ineligible for rollover. Only amounts being distributed in enforcement of 401(a)(9) can automatically be characterized as ineligible for rollover by reason of their being required by the Code.

 

A plan can require (or permit) minimum distributions that are not intended to reflect the requirements of 401(a)(9). For example, a plan can offer or require a series of payments that are to be determined in an identical manner to the Code’s formula for determining RMDs (i.e., over life expectancy). But in that case, if such amount is being paid prior to the Code-required Required Beginning Date (RBD), such series of payments are not Code-required RMDs until the participant reaches the Code-described RBD (which should be stated in the plan).

 

In response to what I view as having been a digression, if amounts are being paid in addition to true RMDs, then such amounts are not triggers for anything other than bigger debits to the participants account balance and the possibility of making a rollover. For example, if the plan’s only provisions relating to age 70½ are those that are intended to reflect the post-SBJPA provisions of 401(a)(9), then there is no such thing as an in-service distribution to a non-owner that exceeds the RMD unless you characterize the entire amount of such distribution as being in excess of the $0 RMD. There can be an amount in excess of a Code-required RMD only if there is a Code-required RMD, and there can be a Code-required RMD only if the participant is described by the plan’s provisions enforcing the Code’s post-SBJPA definition of the RBD. The Code’s post-SBJPA RBD is never triggered by the presence of a distribution (voluntary or otherwise), but only by the combination of age, employment status, and ownership status.

 

Thus, for example, participants can never elect to receive true RMDs early. They might be permitted or required to receive amounts that are determined in exactly the same manner as RMDs, and perhaps the feature is tied to attainment of age 70½, but until the participant has attained the “true” RBD for purpose of the plan’s enforcement of 401(a)(9) (which does not occur until severance from service for non-owners under plans with post-SBJPA provisions), such amounts are not RMDs within the meaning of IRC 401(a)((9) because they are not yet required by 401(a)(9), and the record keeper must account for “pretend” RMDs differently than for “true” RMDs. No plan is required to have annual required minimum distributions other than those required by the Code, and many plans do not. It is worth examining each plan to see how it is constructed.

 

In summary, a document could have a provision that triggers something that can be characterized as being "required minimum distributions," but to answer the original question as I understand it, such required minimum distributions would not be the same thing as 401(a)(9) RMDs because the Code does not use the fact that a distribution has been made as a trigger for determining a participant’s RBD. The Code’s RMD requirements use the Code’s definition of RBD, and a non-owner will not reach that RBD until they stop working as an employee of the employer sponsoring the plan if that plan uses the post-SBJPA definition of RBD. The former ("minimum required distributions) is typically going to be controlled by an AA option, while the latter (401(a)(9) RMDs) will hopefully be found in every plan and hopefully will explicitly refer to IRC 401(a)(9) and will have the appropriate RBD for Code purposes. The plan administrator needs to correctly identify which type of required minimum distribution is being paid.

 

I could understand a plan having a distribution serve as a trigger for ordinary required minimum distributions in order to prevent numerous ad-hoc distributions being requested while in-service (or even after termination). If amending a plan with ad-hoc distributions to provide for such a trigger, beware of IRC 411(d)(6).

 

[asdfgdfa]

On 8/16/2019 at 8:59 PM, Doc Ument said:

 

I agree with everyone who posits that RMDs are never optional, which is why the R stands for Required.

 

The question is: Is the distribution being characterized as “required” by reference to the plan’s language enforcing the Code or is it “required” by some plan provision other than the plan’s language enforcing the Code?

 

If I understand the original question correctly (the identification of which seems to have become the primary topic), the law does not require RMDs simply because a participant chooses to withdraw funds from a plan after attaining age 70½, though the plan might do so just because its designer thought it was a great idea. The law’s only “trigger” is either age alone (i.e. age 70½ for everyone if the plan still uses the pre-SBJPA provision), or (for a plan that uses the post-SBJPA provision) a combination of age, ownership status, and employment status. Only distributions mandated by the Code are “true” RMDs. For example, such amounts are automatically not eligible for rollover if they are required by the Code. In contrast, if the distributions are required only because the plan is fabricating “pretend” RMDs as described below, then such distributions would not be required by the Code and such amounts could not (for example) therefore be automatically characterized as ineligible for rollover. Only amounts being distributed in enforcement of 401(a)(9) can automatically be characterized as ineligible for rollover by reason of their being required by the Code.

 

A plan can require (or permit) minimum distributions that are not intended to reflect the requirements of 401(a)(9). For example, a plan can offer or require a series of payments that are to be determined in an identical manner to the Code’s formula for determining RMDs (i.e., over life expectancy). But in that case, if such amount is being paid prior to the Code-required Required Beginning Date (RBD), such series of payments are not Code-required RMDs until the participant reaches the Code-described RBD (which should be stated in the plan).

 

In response to what I view as having been a digression, if amounts are being paid in addition to true RMDs, then such amounts are not triggers for anything other than bigger debits to the participants account balance and the possibility of making a rollover. For example, if the plan’s only provisions relating to age 70½ are those that are intended to reflect the post-SBJPA provisions of 401(a)(9), then there is no such thing as an in-service distribution to a non-owner that exceeds the RMD unless you characterize the entire amount of such distribution as being in excess of the $0 RMD. There can be an amount in excess of a Code-required RMD only if there is a Code-required RMD, and there can be a Code-required RMD only if the participant is described by the plan’s provisions enforcing the Code’s post-SBJPA definition of the RBD. The Code’s post-SBJPA RBD is never triggered by the presence of a distribution (voluntary or otherwise), but only by the combination of age, employment status, and ownership status.

 

Thus, for example, participants can never elect to receive true RMDs early. They might be permitted or required to receive amounts that are determined in exactly the same manner as RMDs, and perhaps the feature is tied to attainment of age 70½, but until the participant has attained the “true” RBD for purpose of the plan’s enforcement of 401(a)(9) (which does not occur until severance from service for non-owners under plans with post-SBJPA provisions), such amounts are not RMDs within the meaning of IRC 401(a)((9) because they are not yet required by 401(a)(9), and the record keeper must account for “pretend” RMDs differently than for “true” RMDs. No plan is required to have annual required minimum distributions other than those required by the Code, and many plans do not. It is worth examining each plan to see how it is constructed.

 

In summary, a document could have a provision that triggers something that can be characterized as being "required minimum distributions," but to answer the original question as I understand it, such required minimum distributions would not be the same thing as 401(a)(9) RMDs because the Code does not use the fact that a distribution has been made as a trigger for determining a participant’s RBD. The Code’s RMD requirements use the Code’s definition of RBD, and a non-owner will not reach that RBD until they stop working as an employee of the employer sponsoring the plan if that plan uses the post-SBJPA definition of RBD. The former ("minimum required distributions) is typically going to be controlled by an AA option, while the latter (401(a)(9) RMDs) will hopefully be found in every plan and hopefully will explicitly refer to IRC 401(a)(9) and will have the appropriate RBD for Code purposes. The plan administrator needs to correctly identify which type of required minimum distribution is being paid.

 

I could understand a plan having a distribution serve as a trigger for ordinary required minimum distributions in order to prevent numerous ad-hoc distributions being requested while in-service (or even after termination). If amending a plan with ad-hoc distributions to provide for such a trigger, beware of IRC 411(d)(6).

 

Thanks for your reply, this has cleared up a number of things for me. Awesome post!