Possible Change to RMDs? [Executive order]

[Beehave]

Saw this on Yahoo Finance page, from Wall Street Journal:

WASHINGTON—President Trump is expected to sign an executive order Friday directing the government to review rules requiring retirees to start taking annual withdrawals from retirement funds after they turn 70 ½ and to consider making it easier for small businesses to offer employees 401(k) plans.

The action, expected to be signed during a ceremony in Charlotte, N.C., ahead of the Labor Day weekend, was billed by the White House as a push to better prepare workers for retirement.

As part of the initiative, the Treasury Department would review the rules on required minimum distributions from retirement plans to see if investors can keep more money for a longer time in 401(k)s, individual retirement accounts and other tax-sheltered savings plans. If successful, it could allow retirees to spread retirement savings over a longer period.

[SquawkIdent]

Interesting, thanks.

[Alan S.]

Quite possible that the RMD change would simply entail a more current annuity mortality table. Currently, RMDs are based on the 2000 annuity mortality table (blended 50-50 male/female) projected to reflect mortality improvements through 2003. Therefore, we are about due for an update anyway to reflect improvements over the last 15 years. However, the RBD currently used is arbitrary and could be delayed separately or in conjunction with new mortality tables. Don't expect any relief for 2018 as custodians likely object to processing of roll backs.

Of course, this could also trigger another look by Congress at the formerly proposed 5 year rule for non spouse beneficiaries. Such inherited IRAs are typically no longer viewed as retirement accounts, but rather as savings for the beneficiaries which should be taxed much quicker.

[LadyGeek]

The order has been signed. Here's additional information on Executive Orders: What is an Executive Order?, from the American Bar Association.

Both executive orders and proclamations have the force of law, much like regulations issued by federal agencies, so they are codified under Title 3 of the Code of Federal Regulations, which is the formal collection of all of the rules and regulations issued by the executive branch and other federal agencies.

...
Executive orders are not legislation; they require no approval from Congress, and Congress cannot simply overturn them. Congress may pass legislation that might make it difficult, or even impossible, to carry out the order, such as removing funding. Only a sitting U.S. President may overturn an existing executive order by issuing another executive order to that effect.

I don't see anything posted yet on an "official site", such as the Federal Register. - Update: see below.

As a reminder, opinions of the political process are off-topic.

[Artsdoctor]

https://www.whitehouse.gov/presidential ... y-america/

[trueblueky]

Here it is:
https://www.whitehouse.gov/presidential ... y-america/

Lots of examine this and study that. This order doesn't change anything, yet. DOL will conduct the studies and post draft changes for comment at some point.

Updating the life tables is a good idea. Updating them annually may be more trouble than it's worth -- maybe state that they will be updated every x (10?) years rather than willy nilly.

EDIT: I see artsdoctor posted the link while I was typing mine.

[Spirit Rider]

Well, the current RMD tables were for 2003 and the previous we're in 1987. So as Alan stated we are about due for a change.

I vote for reducing the penalty-free distributions by five years to 54 1/2 and increasing the RMD age by five years to 75 1/2.

[NoHeat]

I’d be happy if they just got rid of the 1/2-year thing. And moreover, the April after the 1/2-year thing. That’s so confusing.

[RetiredCSProf]

This proposed change to RMDs was reported on the "Nightly Business News" this evening.

This was reported along with a proposal to reduce capital gains by inflation. That is, to add a factor for inflation to the cost basis of property sold.

[jcar]

The possible change on capital gains tax due to inflation could be quite interesting for real estate

[willthrill81]

I vote for reducing the penalty-free distributions by five years to 54 1/2 and increasing the RMD age by five years to 75 1/2.

I'd be all for that!

But I'm not holding my breath either.

[LadyGeek]

Note that discussion of possible capital gains taxes adjust for inflation is off-topic as "proposed legislation". Here's why: Political comments and proposed tax plan remain off-topic

In summary, the point of the policy is to (1) eliminate contentious disagreements that result from these discussions and (2) keep investors from making bad decisions. Proposed regulations change many times between the time they're introduced and signed into law.

[tfb]

Please note this can only affect the tables. The 70-1/2 required start date is hardcoded in law. Changing that requires new legislation.

Within 180 days of the date of this order, the Secretary of the Treasury shall, consistent with applicable law and the policy set forth in section 1 of this order, examine the life expectancy and distribution period tables in the regulations on required minimum distributions from retirement plans (67 Fed. Reg. 18988) and determine whether they should be updated to reflect current mortality data and whether such updates should be made annually or on another periodic basis.

[daveydoo]

... If successful, it could allow retirees to spread retirement savings over a longer period...

No it wouldn't. It would allow retirees to spread their retirement savings over a shorter period (starting when they are older). The math is not tricky. It would allow wealthy people who do not need to take RMDs to benefit from the longer accumulation phase and then pass the money down to their heirs.

There's one born every minute.

[JoMoney]

Huh.. I wonder how updating the mortality tables might impact people using them for 72(t) SEPP distributions prior to age 59.5.
If someone had already begone taking SEPP distributions using the RMD method, and the tables were changed, would the IRS require they stick to the original distribution table or the updated one?

[cherijoh]

... If successful, it could allow retirees to spread retirement savings over a longer period...

No it wouldn't. It would allow retirees to spread their retirement savings over a shorter period (starting when they are older). The math is not tricky. It would allow wealthy people who do not need to take RMDs to benefit from the longer accumulation phase and then pass the money down to their heirs.

There's one born every minute.

It sounds like retirees would still be starting RMDs at 70.5. But if the adjusted life expectancy tables add a couple of years to your actuarial life span, the divisors would go up and the % of your account that you have to take each year would go down. So I think Beehave is correct about spreading out distributions.

[Strider]

[Thread merged into here, see below. --admin LadyGeek]

As reported in the WSJ and on CNBC, the president has issued an executive order directing that the IRS table for RMDs be re-calculated to account for current life expectancy. My question does not refer directly to that executive order.

Rather, does anyone know how recently the current RMD table was revised? Does it reflect current life expectancy, or alternatively does it reflect life expectancy as of some years ago?

Also, because there has been a widely-reported reduction in American life expectancy due to deaths from opiod use, might this recalculation actually require future RMDs to be larger than currently required?

Strider

[The Wizard]

An actuary would look at the DISTRIBUTION of opioid deaths with age.
It's entirely possible that the bulk of them are people younger than age 70 and thus have little bearing on the conditional life expectancy of people age 70 and beyond...

[FiveK]

According to this page, the latest IRS rules and life expectancy tables...were finalized on April 16th, 2002.

[Strider]

To what extent have life-span projections for the above-70 demographic changed in the last 16 years (since the last revision of the RMD table)? If a lot, this may result in a significant change in the RMD table.

Strider

[Phineas J. Whoopee]

As reported in the WSJ and on CNBC, the president has issued an executive order directing that the IRS table for RMDs be re-calculated to account for current life expectancy. My question does not refer directly to that executive order.
...

Recognizing that your question is not about the executive order, and not knowing what the Journal and CNBC reported, the order is a direction to the Treasury department to review its rules, not to change them. The outcome may or may not eventually lead to a reduction of RMDs.

PJW

[Strider]

Actually, a recalculation of the RMD table may cut either of two ways.

If life expectancy for the over-70 demographic has increased since 2002 (the date of the last revision of the RMD table), then the RMD distribution period would increase in a revised table (lower RMDs).

However, if life expectancy for the over-70 demographic has decreased since 2002 (due perhaps to deaths from opiod use), then the RMD distribution period would decrease in a revised table (higher RMDs).

I don't know how to access the life expectancy data that the Department of the Treasury might use.

Strider

[LadyGeek]

FYI - The executive order is discussed here: Possible Change to RMDs? [Executive order]

Update: See below.

[LadyGeek]

I merged Strider's thread into here. While Strider is indeed asking a different question, both threads are now discussing the RMD tables.

[celia]

I’d be happy if they just got rid of the 1/2-year thing. And moreover, the April after the 1/2-year thing. That’s so confusing.

What would you propose instead?

I would think this is to take into account those who turn 70.5 (or 70) on December 30 or 31. How are they supposed to take the first RMD in the same year if they want to wait for their birthday or half-birthday that year? What if it is a holiday weekend? What if they are traveling? What if they want it to be a Qualified Charitable Donation and they need to get it to the charity and also have the charity cash the check in the year of donation. There just aren't enough days between December 31 (half) birthdays and the end of the year to make much happen. [sic--I am aware of what I am saying in the last sentence]

Maybe you would propose:

The Required Beginning Date (RBD) Year (RBY) is the year in which the taxpayer turns 71.

This would allow everyone born the same year to take their RMD anytime they want that year. And they could allow the first RMD, or part of it, to be used for a QCD, even if the taxpayer turned 70 on the previous December 31 and wanted to donate before July 1.

[The Wizard]

I’d be happy if they just got rid of the 1/2-year thing. And moreover, the April after the 1/2-year thing. That’s so confusing.

What would you propose instead?

I would think this is to take into account those who turn 70.5 (or 70) on December 30 or 31. How are they supposed to take the first RMD in the same year if they want to wait for their birthday or half-birthday that year? What if it is a holiday weekend? What if they are traveling? What if they want it to be a Qualified Charitable Donation and they need to get it to the charity and also have the charity cash the check in the year of donation. There just aren't enough days between December 31 (half) birthdays and the end of the year to make much happen. [sic--I am aware of what I am saying in the last sentence]

You've been around long enough to know that RMDs are due starting January 1st of the YEAR you turn 70-1/2, or whatever new age might be legislated.
Am I right?

[FiveK]

You've been around long enough to know that RMDs are due starting January 1st of the YEAR you turn 70-1/2, or whatever new age might be legislated.
Am I right?

Not according to the IRS.

See When must I receive my required minimum distribution from my IRA?:

You must take your first required minimum distribution for the year in which you turn age 70½. However, the first payment can be delayed until April 1 of the year following the year in which you turn 70½. For all subsequent years, including the year in which you were paid the first RMD by April 1, you must take the RMD by December 31 of the year.

[The Wizard]

You've been around long enough to know that RMDs are due starting January 1st of the YEAR you turn 70-1/2, or whatever new age might be legislated.
Am I right?

Not according to the IRS.

See When must I receive my required minimum distribution from my IRA?:

You must take your first required minimum distribution for the year in which you turn age 70½. However, the first payment can be delayed until April 1 of the year following the year in which you turn 70½. For all subsequent years, including the year in which you were paid the first RMD by April 1, you must take the RMD by December 31 of the year.

You are correct.
So that allows even more time for someone to take their required amount...

[celia]

You are correct.
So that allows even more time for someone to take their required amount...

Even if you delay until the following April for your first RMD, the second one is also due in the same year. You are only delaying the first one a few months.

[blackcat allie]

Without quoting from order directly, it appears that fiduciary is now on the line in retirement accounts (for the purpose of expense cutting and thereby increasing ease of use by employers). Did anyone else get that impression?

[Phineas J. Whoopee]

Without quoting from order directly, it appears that fiduciary is now on the line in retirement accounts (for the purpose of expense cutting and thereby increasing ease of use by employers). Did anyone else get that impression?

I didn't get that impression. The order directs Treasury and Labor to review some of their rules, not to change them. It may or may not be the case that eventually some changes will occur.

PJW

[jclear]

To what extent have life-span projections for the above-70 demographic changed in the last 16 years (since the last revision of the RMD table)? If a lot, this may result in a significant change in the RMD table.

Strider

Apparently the government's 2002 regulations were based on a 2000 commissioned actuarial report.
https://www.soa.org/experience-studies/ ... ty-tables/
The thing that gets me is that the RMD were set based off these, but not just 1/expectedremaininglife. Does anyone working there even remember on what basis they calculated the RMD tables? They can look at their records, but new minds new thinking. If their calculations included inflation, interest rates, and so on, then things have changed.

[ralph124cf]

I’d be happy if they just got rid of the 1/2-year thing. And moreover, the April after the 1/2-year thing. That’s so confusing.

What would you propose instead?

I would think this is to take into account those who turn 70.5 (or 70) on December 30 or 31. How are they supposed to take the first RMD in the same year if they want to wait for their birthday or half-birthday that year? What if it is a holiday weekend? What if they are traveling? What if they want it to be a Qualified Charitable Donation and they need to get it to the charity and also have the charity cash the check in the year of donation. There just aren't enough days between December 31 (half) birthdays and the end of the year to make much happen. [sic--I am aware of what I am saying in the last sentence]

You've been around long enough to know that RMDs are due starting January 1st of the YEAR you turn 70-1/2, or whatever new age might be legislated.
Am I right?

The law seems quite clear: If you turn 70.5 on December 31, 2018, you may take the RMD at any time during 2018, however, for a QCD, that can only be done AFTER you turn 70.5. The law does not have to make sense.

Ralph

[jeffyscott]

To what extent have life-span projections for the above-70 demographic changed in the last 16 years (since the last revision of the RMD table)? If a lot, this may result in a significant change in the RMD table.

Strider

Apparently the government's 2002 regulations were based on a 2000 commissioned actuarial report.
https://www.soa.org/experience-studies/ ... ty-tables/
The thing that gets me is that the RMD were set based off these, but not just 1/expectedremaininglife. Does anyone working there even remember on what basis they calculated the RMD tables? They can look at their records, but new minds new thinking. If their calculations included inflation, interest rates, and so on, then things have changed.

The IRS Uniform Life Expectancy Table assumes that distributions would extend over two lives: yours and a beneficiary 10 years younger than you, according to something I read at Fidelity. This makes sense, since if your spouse is over 10 years younger, you'd use the Joint Life and Last Survivor Expectancy Table.

[bsteiner]

As others have pointed out, it's been about 15 years since the tables were updated, so it might make sense to update them. If life expectancies have increased, that would reduce the required distributions.

It might be complicated to provide for multiple employer retirement plans, but I think it's worth the effort to see what they can come up with. In that regard, I would be interested in hearing from those who run small businesses as to how burdensome it is for them to set up and maintain a plan for their business.

[NoHeat]

You've been around long enough to know that RMDs are due starting January 1st of the YEAR you turn 70-1/2, or whatever new age might be legislated.
Am I right?

Not according to the IRS.

See When must I receive my required minimum distribution from my IRA?:

You must take your first required minimum distribution for the year in which you turn age 70½. However, the first payment can be delayed until April 1 of the year following the year in which you turn 70½. For all subsequent years, including the year in which you were paid the first RMD by April 1, you must take the RMD by December 31 of the year.

This discussion illustrates my point that the present rules are unnecessarily confusing. Average retirees don't have much chance of always getting it right, on their own. That's why I'm hoping the rules get simpler, or at least not any more complicated, if they get revised in the near future.

[Spirit Rider]

You've been around long enough to know that RMDs are due starting January 1st of the YEAR you turn 70-1/2, or whatever new age might be legislated.
Am I right?

Not according to the IRS.

See When must I receive my required minimum distribution from my IRA?:

You must take your first required minimum distribution for the year in which you turn age 70½. However, the first payment can be delayed until April 1 of the year following the year in which you turn 70½. For all subsequent years, including the year in which you were paid the first RMD by April 1, you must take the RMD by December 31 of the year.

This discussion illustrates my point that the present rules are unnecessarily confusing. Average retirees don't have much chance of always getting it right, on their own. That's why I'm hoping the rules get simpler, or at least not any more complicated, if they get revised in the near future.

So you would prefer the the 1st year 4/1 rule was not in place???

• The one that allows you until 4/1 following the 1st year RMDs are due in case you didn't know or ran out of time?
• The one that allows you to intentionally defer the 1st year RMD to the next year after you retire when you may be in a lower tax bracket.

So no, the rules are not unnecessarily confusing. They are in fact necessary for the benefit of the taxpayer.

In this case, if you didn't know about the 4/1 following the 1st year rule. You are no worse off than if the rule never existed. However, those who take the time to learn can benefit.

[Strider]

Still another aspect to the current rules of which many are unaware is the ability to split the first (but ONLY the first) RMD between two calendar years. Link: https://www.irahelp.com/forum-post/2371 ... ear-period

Strider

[brennok]

So if you are already taking RMDs would this affect you? Or would you just continue to use the old table? Does anyone know what happened in 2002 when the tables were updated?

[The Wizard]

...In this case, if you didn't know about the 4/1 following the 1st year rule. You are no worse off than if the rule never existed. However, those who take the time to learn can benefit.

As a practical matter, delaying first RMD until 4/1 of following year is only helpful for certain folks who worked full time till 70+ or otherwise had high one-time taxable income in the year they turned 70.5.

You can't do any Roth conversions starting 1/1 of the year you turn 70.5 until you complete your RMD, so that's not a plan.

As for me, I'm going to take monthly withdrawals from tax deferred totalling up to my RMD starting January, 2020.
And those withdrawals look to be a chunk less than what I've been taking since 2013, in lieu of SS...

[JoMoney]

So if you are already taking RMDs would this affect you? Or would you just continue to use the old table? Does anyone know what happened in 2002 when the tables were updated?

Yeah, that was my question too... especially for any early retirees who may be using the tables for 72(t) SEPP withdrawals

[The Wizard]

So if you are already taking RMDs would this affect you? Or would you just continue to use the old table? Does anyone know what happened in 2002 when the tables were updated?

Yeah, that was my question too... especially for any early retirees who may be using the tables for 72(t) SEPP withdrawals

I'm pretty sure you just start using the new table for the yearly divisor henceforth.
That's not so hard to imagine, is it?

[JoMoney]

So if you are already taking RMDs would this affect you? Or would you just continue to use the old table? Does anyone know what happened in 2002 when the tables were updated?

Yeah, that was my question too... especially for any early retirees who may be using the tables for 72(t) SEPP withdrawals

I'm pretty sure you just start using the new table for the yearly divisor henceforth.
That's not so hard to imagine, is it?

It's not hard to imagine, but correct guidance is critical. Incorrect interpretation could cause a massive penalty.
The current guidance is that once you've selected a life expectancy table it can not be changed. It's implied that they're saying you can't choose to switch from a Single Life Table to a Uniform Life table (or other table)... but it's not explicit other than saying the table can't be changed.

[jclear]

Apparently the government's 2002 regulations were based on a 2000 commissioned actuarial report.
https://www.soa.org/experience-studies/ ... ty-tables/
The thing that gets me is that the RMD were set based off these, but not just 1/expectedremaininglife. Does anyone working there even remember on what basis they calculated the RMD tables? They can look at their records, but new minds new thinking. If their calculations included inflation, interest rates, and so on, then things have changed.

The IRS Uniform Life Expectancy Table assumes that distributions would extend over two lives: yours and a beneficiary 10 years younger than you, according to something I read at Fidelity. This makes sense, since if your spouse is over 10 years younger, you'd use the Joint Life and Last Survivor Expectancy Table.

There's a separate table for people with their spouse as beneficiary and their spouse is 11 years or more younger than them. The ULET is for the case of a single person, or an account where the spouse is 10 or fewer years younger (or is older), or where the spouse is not the sole beneficiary. The 11+ year difference table is a simple joint survivor life expectancy table. Comparing the tables, it looks like the person using ULET is given the consideration of having a spousal beneficiary exactly 10 years younger than themselves, regardless of the exact situation. So, yes, 590B confirms what Fidelity said.

[Jablean]

For those of us dealing with RMDs on inherited IRAs wonder if they'll let us recalculate as our factors are set on date of death.

[Alan S.]

For those of us dealing with RMDs on inherited IRAs wonder if they'll let us recalculate as our factors are set on date of death.

When the 2002 tables were released, the divisor for inherited IRAs (single life table) was then determined by applying the 1.0 annual divisor reduction to the revised first year divisor shown in the new table. Therefore, inherited IRA RMD were reduced as well as owner's RMDs. Most likely, this methodology would be applied again if the tables are revised.


There is a lack of transparency to the entire process of RMD table revision and how it is triggered. Would be nice if more transparency was provided, if not specified in IRS Regs.

[ralph124cf]

A poster above mentioned the ability to split the RMD for the first year after you turned 70.5. There was also a question about QCDs for somebody who turned 70.5 on December 31st.

As far as I can see you can just double up the QCD the second year and everything is a wash.

For example, suppose you have an RMD of $100K both years, and you want to give a QCD of $15K each year. Simply take 85K as your partial RMD the first year, then the second year first give a QCD of $30K, and then take the remaining $85K to complete the second year RMD.


Any problems with this scenario?

Ralph

[Spirit Rider]

Any problems with this scenario?

I don't see any.

Of course, it will only reduce the AGI in the following year, but by double.