For years, and especially the last few months, there have been numerous headlines trumpeting “The Death of the Stretch IRA!” Take a look:
Ed Slott in InvestmentNews
Lewis Braham in Barrons:
Even CNBC got in on the act:
There's only one problem. The Stretch IRA didn't die. It just changed. In one respect, it got better. In another respect, it got worse. (Or if you're Congress, the IRS, or anyone who thinks government should have more money to do more stuff, in one respect it got worse and in another respect, it got better.) Let me explain.
Change # 1 No RMDs Required for Inherited IRAs
The first change, which everyone seems to gloss over, is the elimination of a requirement to take an RMD from the inherited IRA every year. You don't have to take a single withdrawal from that sucker until year 10. That is a substantial advantage.
Consider someone who leaves their Roth IRA to their 80-year-old little sister. That woman would be required to withdraw 5.3% of that IRA this year, 5.6% next year, and so on until age 90, when she would be required to take out 8.8%. How much of a difference would that make for her to not have to take anything out until year 10?Well, let's assume it's a $100,000 Roth IRA growing at a miraculously even 8% a year (in truth, variable returns would make this effect worse). If she didn't take anything out for 10 years, it would be worth $215,893. With those withdrawals, it would only be worth $108,021, half as much!
Now, to be fair, you would have taken $77,301 out of the Roth IRA, and if you had invested that in some reasonable way (let's say it earns 6% after-tax) it would have grown to $99,043, for a total of $207,065. So you end up with $8,828 (4.3%) more. That's a nice little kicker for tax-free growth. But you also eliminate the hassle of having to remember to take the RMDs, any possible penalty for forgetting to do so (50% of the amount you should have taken), and the costs of reinvesting the money.
This is a nice benefit. This is the New Stretch IRA. If your heir will just sit on that IRA, it is likely to at least double in size over the next decade. It's not the same stretch IRA we used to have, but it still beats no stretch IRA at all.
Change # 2 You Have To Withdraw All the Money By Year 10
Here is the Stretch IRA change all the headlines above are bemoaning. Instead of potentially being able to leave your IRA to your great grandkid and having them stretch it over the next 80 years, that kiddo has to withdraw all the money in year 10. I just find it hilarious that people are whining about this. It truly demonstrates just how much the human psyche hates to lose something, even if it is something almost nobody actually ever had. Most IRAs have never even been inherited yet because most people who started them haven't died. Traditional IRAs have only been around since 1974, 401(k)s didn't exist before 1978, Roth IRAs didn't show up until 1997, and Roth 401(k)s didn't start until 2006. 401(k)s became popular in the early 1980s, but the employee contribution didn't even hit $10K until 1998, just in time for everyone to lose half their money anyway in the tech meltdown.
So points # 1 and # 2 are that very few people have ever actually inherited an IRA and that IRA was likely pretty small anyway. But point # 3 is the real killer: Most heirs were never going to maximally stretch an IRA anyway. I mean, people are people. You give somebody $50K, $100K, even $500K or a million and what are they going to do with it? You really think they're only going to take out 1, 2, 3% a year? I've got a bridge to sell you. You might think your kids are all going to be high earners and financial whizzes, but let's not kid ourselves. They're going to buy boats, drive Teslas, buy a bigger house than they could otherwise afford, and send their kids to private preschools and colleges.
I think they'd be doing pretty good if that inherited IRA lasted longer than 10 years anyway! And that's assuming they can invest their way out of a paper bag. Maybe they panic in the first bear market they see and sell low, losing half its value before spending the rest on a brand new Tesla Z with incrediludicrous mode and 6 wheel drive.
Point # 4? You probably weren't going to leave that thing to your great grandkid. Probably not even your grandkid. You were probably going to leave it to your kid, who is maybe 25 years younger than you. If you die at 95 they're already 70, their RMDs will start at 3.6%, and at most they would be able to stretch it 20 or 30 years.
Enough ranting though. Only being able to stretch an IRA for one decade instead of 2-7 decades truly is a loss for those precious few who built a huge IRA, didn't spend it, and left it to a young, financially savvy heir. But wait, there are exceptions to this rule.
The Exceptions
Of course, this new change isn't universal. There are some people who can still stretch an IRA indefinitely. These “Eligible Designated Beneficiaries” include:
- Your spouse
- Any disabled heir
- Individuals less than 10 years younger than the deceased
- Minors (but only until they turn 18).
You may see some lists that include “chronically ill” as a definition, but the definition of chronic illness is basically the same as being disabled so I didn't include it as a separate category.
What You Should Do Differently Now
So here's where the rubber meets the road. You know the law changed. You know what the changes are. But what should you do about it?
# 1 Designate Eligible Beneficiaries
You will likely have lots of heirs, why not preferentially leave the IRAs and Roth IRAs to designated beneficiaries? I mean, most will likely leave it to their spouse (who will probably still usually roll it into their own IRA rather than taking it as an inherited IRA anyway), but when the second spouse dies, then you've got some decision making to do. You can still leave it to a great grandkid. They might get almost 2 decades of stretching out of it. You can also leave it to your sibling or other elderly person. If you have one disabled kid and one healthy one, why not leave the IRA to the disabled one and the taxable account to the healthy one? The exceptions are definitely something to be aware of when doing your estate planning and designated beneficiaries.
# 2 Fix Your Trust
If you designated a trust as the beneficiary of your IRA, you probably need to go see your estate planning attorney again. If you have a conduit trust as the IRA beneficiary that specifies that only the RMDs are distributed each year, that heir isn't going to get squat for 9 years and then is going to get a huge bolus in year 10. Probably not what you were hoping for. If you have a discretionary trust, all that money is going to be sitting in there after 10 years getting taxed at the high trust rates. The bottom line is that you probably want to change the directions for the trustee to follow so that your money doesn't end up with the IRS instead of your heir.
# 3 Beg For the Roth IRA
If your parents are passing out their IRAs, ask if you can be the one who gets the Roth IRA. Not only are you likely to get more after-tax money (since many of your siblings may not recognize that a traditional IRA isn't worth as much as a Roth IRA), but you can just leave it in there for a decade and then take it all out at once. Much easier.
# 4 Plan Your Withdrawals Carefully
If you do end up inheriting a traditional IRA, all is not lost. You have 10 years to get that money out. Obviously you want to leave it in there as long as possible to facilitate tax-protected growth (and asset protection), but you also don't want to pay any more taxes than you otherwise would have to.
- If you're in the top tax bracket and always will be, well, just leave it there for all 10 years and take it out at the end.
- If it's a tiny IRA and won't even get you near the next tax bracket, again, leave it there for all 10 years.
- If it is a large IRA, then you need to plan a little bit more carefully. Let's say you're married filing jointly with a taxable income of $250K and its' a $200K IRA, likely to be $400K in 10 years. You are currently in the 24% tax bracket and can take an additional $77K a year in income without having to pay 32% on it. $400,000/77,000 = about 5 years. So you should probably start taking withdrawals up to the top of the 24% tax bracket after about 5 years. Obviously, if your income or tax rates change, you may have to adjust the plan.
- If it is a really large IRA, you may need a more extreme plan. Maybe you take a year sabbatical and pull $300K from the IRA all at once. Or maybe you plan your retirement date around the last couple of years of the 10 year period and spend the inherited IRA first.
# 5 Consider Doing More Roth Conversions
If you expect your heir to be in a high tax bracket/peak earnings years for all ten years when they can withdraw from your IRA, consider doing more Roth conversions so that they can stretch it out for all 10 years.
The Stretch IRA isn't gone, but is different (at least for most heirs). It's important to understand how the New Stretch IRA works and take advantage of it.
What do you think? What changes will you make in your financial or estate plan as a result of these changes? Do you have an inherited IRA? How long have you been stretching it? Comment below!
“few people have ever actually inherited an IRA and that IRA was likely pretty small anyway. But point # 3 is the real killer: Most heirs were never going to maximally stretch an IRA anyway. I mean, people are people. You give somebody $50K, $100K, even $500K or a million and what are they going to do with it? You really think they’re only going to take out 1, 2, 3% a year? I’ve got a bridge to sell you.”
That’s a bold assertion provided with no citations.
My dad (a middle class family doctor) died 17 years ago at age 63 with $1.5M in his profit sharing account. That converted into an IRA and I and my four siblings have been taking RMDs ever since. My account has grown to about twice what it started.
I suspect the majority of your readership is in a similar situation with profit sharing accounts that can be rolled into an IRA.
There’s a lot of misconceptions about IRAs but this assertion in particular strikes me as plain wrong for your readership.
You have correctly identified the counter argument. Just because most people won’t stretch it out for decades, doesn’t mean some won’t.
But now you can’t. So deal with it.
As always, I enjoy your posts and have learned so much from them. My only comment is that you might actually be surprised at how carefully and frugally people spend money that has been inherited. Don’t just assume that an heir will go on a reckless spending spree.
He nailed it when it comes to my family!
My mom is in her 80’s and will likely leave over a million dollars (after she and my dad scrimped and saved for decades) to my brother, sister and I , and 4 grandkids. My brother and sister will spend that money so fast it will make your head spin (as evidenced by how they have spent their money their whole lives including the recent stimulus checks-one on a $7000 adjustable bed-while already in debt). My siblings have always been in debt and can’t seem to figure out basic finance, and do NOT want unsolicited advice from me. They have never grasped the concept of saving- what good is holding onto money for later when you need it now? Their kids are more responsible with money. I am trying to help mom plan how she wants the money distributed because she knows they will withdraw any money they inherit and spend it the first year (most is in a traditional IRA, some taxable).
So maybe not applicable to everyone but rings very true for me!
You have correctly identified the counter argument. Just because most people won’t stretch it out for decades, doesn’t mean some won’t. Hopefully many of my readers are in that category.
But again….now you can’t. So deal with it.
Interesting article as I’m drafting my estate plan now with my attorney. He has included language about the SECURE Act in the documents so it’s good that he’s staying up with the literature and recent events. However your article, specifically point #2 has got me thinking. For the retirement accounts, his recommendation is to have the surviving spouse as primary beneficiary but the contingent beneficiary as my children’s IRA beneficiary trust (kids are minors). I’m curious how #2 applies in that scenario.
It applies to the trust, but not the spouse. So if the effective beneficiary is the contingent beneficiary at your death, then the stretch will be limited to 10 years.
What am I missing here?
Most chronically ill people are not disabled. I specialized in occupational and environmental medicine and a large part of my job was keeping “chronically ill” people in the workplace.
In my own situation, my son has cystic fibrosis. He will likely be able to continue working into middle-age, with his chronic illness. But he may not be able to work full-time. The same is true of someone with multiple sclerosis or a variety of other conditions.
the distinction would make a big difference for him because he’s approximately 45 years younger than I am.
also, he may not be disabled when I die but he more likely than not will become disabled 15-20 years after I die. But he will be defined as having a chronic illness when I die as, it’s an incurable genetic disorder.
This scenario will be much more common than many may think.
Secondly, if you die at 90 and your child was 70 years old, they would not be considered “disabled” because they would be eligible for regular Social Security, not SSDI (Soc Sec Disability Insurance)… Correct? BUT your 70 child could be chronically ill, in fact he/she likely would have at least one chronic illness. I think it’s an intentional loophole in the law, that was placed there so you could leave your IRA to your child in most instances without the 10 year rule.
Please correct me if I’m wrong. You’re the expertand I’m just trying to learn.
An excellent point and potentially a very large loophole. Who doesn’t have some sort of chronic illness?
The SECURE Act modified the Federal LTC definition for “chronically ill”, as it relates to stretching RMDs, from certification by a licensed health care practitioner as being unable to perform two or more ADLs for 90 days to “an indefinite one which is reasonably expected to be lengthly in nature”. Health and safety issues related to severe cognitive impairment would also apply. To meet the “Eligible Designated Beneficiary” requirement, the chronically ill beneficiary must be significantly impaired, not only be diagnosed with a chronic illness. Since the modification is not a hard timeline, I think it likely that the words “reasonably” and ” lengthly” will be litigated.
Adding two ADLs to the list certainly dramatically reduces the number of people who will qualify for that exception.
Thanks for another helpful post! My 76 year old mother has a traditional IRA that’s just over a million dollars. She has another 500K or so in taxable family trust accounts that she and my dad put together before he died. I’m not 100% sure how those (survivor’s and decedent’s) trust accounts work, but I’m guessing you’d suggest that she do some Roth conversions? I’m the only kid and main heir so I’m the one who will likely be affected down the line. Thanks again!
I don’t know enough to recommend Roth conversions or not. Why would she want to do a Roth conversion? I’m not seeing a reason. Now YOU might want her to do some Roth conversions, in which case you should pay for them since she only has $1.5M.
As far as how the trusts work, read the trust document. If you still don’t understand it, consult with the attorney. Every trust is a little different. It’s likely just a revocable trust though to facilitate estate planning.
Interesting post. I don’t have an inheritance to worry about but hope to leave to my heirs. Based on your post, it seems the best strategy may be to max out back door Roths as we would do anyway and then make sensible Roth conversions in lower tax bracket years in retirement accounts that will be rolled to IRAs upon inheritance? Is there something I’m not thinking about? Still early in my financial education. Thanks!
That all sounds reasonable.
I am still doing the calculus on all of this. Don’t have to decide this year since it’s a high tax bracket year, but next year might need to choose whether I should fill up the 24% bracket with conversions. However, considering that converting $1 Traditional to $1 Roth costs 24 to 36 or more cents from other funds, am I certain I’d rather do that than let my heirs bump the basis of my 40 year old mutual funds down to $0 when they inherit? At some point of the variables- our age at death, how much we actually end up leaving, changing tax laws, etc., that could hold a lot more money back from the IRS. I’m considering ensuring Roth to Trad (including 401K equivalents?) is 50-50, but then WCI did an article that woke me up as to whether leaving an inheritance of taxable funds or Roth IRA saves more in taxes.
Others pointed out the valid counterpoint of loss of a very good generational wealth vehicle. It wasn’t really meant to be there and it was closed, so we just need to optimize the new plan.
The concern with this blog post though is that it he points of the argument against the counterpoint during the rant. It makes out that your beneficiaries are poor financial folk, don’t have much in the IRA given small horizon history, or simply that people take the money and run.
BUT, the recommendations to optimize do exactly what the presumptions of the beneficiaries are NOT.
Keeping the iIRA ten years clearly is the best option now. But if you take the premise of the beneficiaries to be true, the recommendations fall short anyways.
That’s the problem. I get it. Folk are upset in the loss of a very nice generational wealth transfer tool and are grieving. Let them grieve this year and hold the rant . In short, deal with it.
I think the recommendation is to avoid touching the inherited IRA for 10 years whenever possible, no matter whether the decedent or heir are rich or poor. I just don’t think most who inherit one will do so, whether rich or poor.
Is the elimination of RMD for inherited IRA the same for traditional and roth?
Yes, if the beneficiary does not qualify as an “Eligible Designated Beneficiary”, then distributions after the death of the IRA owner must be made by the end of the tenth calendar year, following the year of death, for both traditional and ROTH IRAs. It also applies similarly to employer sponsored plans of both persuasions. Exceptions and delays in implementation apply to certain annuities, collective bargaining plans and governmental plans. Greater depth of detail can be found at: https://www.kitces.com/blog/secure-act-stretch-ira-401k-elimination-eligible-designated-beneficiary-retirement-accounts-taxes/
My father is 67 and currently non-dependent on his IRA and probably(hopefully) won’t be for a while. He has not started collecting SSI because of citizenship verification issues(he immigrated here when he was 6 years old and has been paying payroll taxes for nearly 50 years). He did have a job making 30K a year for the last several years but is now currently on unemployment. Would it be a good idea to start making incremental conversions to a Roth starting now in order to empty as much as possible before he hits 72? If so, what percentage would be recommended? Also, the 5 year wait for a Roth withdrawal does not apply to people over the age of 59.5 correct?
Thank you!
Without current financial status info for both you and your father, both your current and projected marginal tax brackets, and a strong sense of your father’s legacy goals, it is not possible to determine whether ROTH conversions are a good idea or not. This might be a good project for an hourly/flat fee financial planner.
As to the two ROTH 5-year rules, rather than summarize what has already been written, I suggest the following article on Michael Kitces’ site:
https://www.kitces.com/blog/understanding-the-two-5-year-rules-for-roth-ira-contributions-and-conversions/
Although it is an older post, I believe it is still accurate.
Likely yes, but without more details, it’s hard to say. In general, you convert up to the top of a tax bracket, not a certain percentage.
There are two 5 year rules with regards to Roth IRAs. I’m not sure which one you’re referring to. You can read more here:
https://www.whitecoatinvestor.com/the-5-year-rule-for-roth-ira-conversions-podcast-146/
That’s a good point. It could be nearly 11 years for many.
Yes.
Excellent post with actionable advice. When considering leaving assets to a disabled person it’s especially important to consult with an estate planning attorney – the last thing you want to do is leave assets to someone who is disabled and that results in them losing benefit eligibility under Medicaid/SSI.
So true. We have adult child and setup an irrevocable special needs trust for him to avoid SSDI loss and clawbacks. At least until we reach retirement age and start pulling SS ourselves and he can be beneficiary at that point and pull on our SS.
Also pensions will come into play for that and that’s where we are going to need a planner to decide on pension elections as well as current IRA laws at that time. Fortunate enough to have these items for next generation and grandkids, but will need to know best efficiency in wealth transfer to their long term benefit and future generations — regardless if they are good stewards or spoiled trust fund kids. .
A few comments:
A correction to consider: Minor children of the decedent are “Eligible Designated Beneficiaries” , all other minors, including grandchildren, great-grandchildren, etc. are generally subject to the 10-year rule. The great-grandchild mentioned in #1 will only get 10 years.
Minor children of the decedent become non-eligible upon attaining the age of majority, which is a state based designation. There a a few states where this is older than 18. Based on interpretation of other Treasury rules, if the beneficiary is taking a specified course of education, they should be able to continue to stretch the IRA till they reach the age of 26.
An important part of analyzing what to do when planning for generational wealth transfer is to consider the beneficiary’s expected marginal tax rates. This can impact whether ROTH conversions, charitable intent, etc. may have a valuable impact.
Good points.
Thank you for the thought-provoking article.
I have a $2.2M Tax-Deferred IRA that I am converting to Roth over the next 5 years before I hit 72. In Tax Year 2020 I’m converting $215k, which (per my CPA) keeps us just inside the 24% marginal Tax Bracket ($326,600, married, filing joint). However, with such a large Tax-Deferred IRA balance, it seems to me that it might make sense to do a Roth conversion that’s *larger* than that $215k – not only in TY 2020 but also in each of these next five years, even though that would that push us up into the 32%, 35% or even 37% bracket. The (simplified) argument is this: it’s better for me to convert it all to Roth and to pay (say, up to) 37% tax on a balance of $2.2M (i.e., pay $814k in taxes), than it is to *not* convert. If I did *not* convert, my heirs would pay (a minimum of) 24% tax on a balance that in 20 years, when I say my farewells, may grow to (say) $8M. They’d pay a staggering $2M in taxes. If my heirs are in a higher tax bracket than 24% the taxes on that $8M would be even more breathtaking.
In summary: better to pay $814k taxes now than $2M in taxes in 20 years, yes?
My day-to-day income and cash flow are good, and will further improve when I start taking Social Security at age 70, in 3 years. I don’t (and won’t) need the Tax-deferred IRAs nor the Roth IRAs, and my legacy goal is to leave it all to my heirs. As needed, I would pay the taxes on the conversions out of savings and Taxable accounts.
Thank you. I’d appreciate any comments.
It is unlikely to converting it all to a Roth IRA is the right move. Your logic is also flawed. It is all about the rates, not the amounts of tax that will later be paid.
Certainly your heirs would prefer you pay the taxes on the money if their inheritance will be the same size, but otherwise, you should strive to have the taxes paid at the lowest possible rate, which might be theirs.
Thank you – I appreciate your response. My goal is to maximize the net amount that my heirs end up with. In order to achieve that, I don’t mind if *I* pay tax now or if *they* pay tax later – whichever makes most sense. Looking at the numbers above, and focusing *not* on the *taxes paid* but instead on *the net amount they actually receive*, it looks like this:
Assumptions (we have to start somewhere – these are rough, simplified numbers, I acknowledge):
1. Portfolio growth: Over a 20 year time horizon, let’s say that the $2.2M Tax-Deferred IRA doubles in value (that’s well below historical average growth rates, so $4.4M is a conservative ending balance projection).
2. Tax rates for future generations will probably increase (huge national debt). But, for argument’s sake, let’s assume they stay the same.
3. My heirs are young (ages 22 & 23), and have well-paying jobs just out of college, and are *today* in the Single, 22% bracket. In 20 years they will be mid-career, hopefully earning well. If we assume conservatively – worst case – that their careers (and incomes) do *not* progress, and that in 20 years they are married filing jointly, that’d put them in the 24% tax bracket.
Scenario 1: Do *NOT* convert any additional, and therefore *stay* within my current 24% tax bracket. Just leave the $2.2M Trad IRA *as is*, and let the heirs pay the taxes. The math: Taxes that *I* pay: Zero. Taxes that *they* pay: they pay 24% on what has grown to become a $4.4M Trad IRA (i.e., they pay $1,056,000), leaving them with a *net* of $3,344,00.
Scenario 2: Full conversion to Roth: over the next five years I convert *all* the $2.2M Trad IRA, which moves me from my current 24% tax bracket to the upper end of the (say) 35% bracket. The math: Taxes that *I* pay: $770,000. Taxes that *they* pay: Zero. Their *net* legacy: $4.4M.
Summary:
Scenario 1: Total taxes paid: $1,056,000 (by the heirs). Net legacy amount after taxes: $3,344,00.
Scenario 2: Total taxes paid: $770,000 (by me). Net legacy amount after taxes: $4.4M.
It seems to me that Scenario 2 (full Roth Conversion) is a no brainer – lower taxes paid, and (more importantly), a higher net legacy at the end of the day. Furthermore, Scenario 2 becomes *even more* of a no brainer if we challenge those three very conservative assumptions: (1) if portfolio growth is higher than stated, (2) if tax rates*do* increase, and (3) if my heirs’ incomes *do* increase, such that they will be in a higher bracket (such as 35%!).
Where is my logic still flawed? What am I missing?
Thank you for your comments.
The logical flaw in your logic is that in scenario two you’re putting more money into the account. Of course scenario two ends up with more money. You need to compare both the $2.2M IRA AND the money that would go toward the taxes versus $2.2M in a Roth IRA. But I think if you compare apples to apples, you’ll go for scenario 1, especially since by the time you keel over and the 10 years run on the stretch, they may not be working at all and will be in a much lower bracket than 35%.
Without checking all your math, the only way you get a huge difference like you are suggesting is for there to be a huge tax difference between the money converted to Roth vs taxes paid on IRA spent. I am not seeing a huge difference there so let’s do a back of the napkin check on your math.
20 years you say $2.2m grows to $4.4 – that is a growth rate of 3.526%, so keep that in mind.
You take 5 of the 20 years to convert $2.2m while still growing and say a current salary of $200k before conversions). The result is roughly $1.5m after paying $860k in tax. That $1.5 million now grows to roughly $2.5 million after 15 years. Less than the IRA of $3.3 million. NOTE: It does not double because you have used up 5 years in doing the conversion.
The problem seems to be you are assuming you pay the conversion tax from a taxable account which you would need to do if you are under age 59.5. Like WCI says you are then adding external money to the Roth out of your “back pocket.”
Thank you, White Coat, and Financial Dave – greatly appreciated.
To WCI’s first point – I think I follow along. With that in mind:
Let’s start with Scenario 2 (Full conversion to Roth: over the next five/six years): Let’s say I decide to do an *additional* conversion (beyond the $215k my CPA said keeps me just within the 24% bracket (gross income up to $326,600, married, filing jointly). In this hypothetical example, I’d be making a conscious choice to go into the upper end of the *32%* bracket (income up to $414,700, married, filing jointly). So the *additional* amount I’d be converting, beyond the original $215k, would be $88,100. To WCI’s point, the $32% tax due on that amount – which I’d pay now from a Taxable account – would be $28,192, and yes, as you say: “you’re putting more money into the [Roth] account.” In this scenario my heirs would pay *zero* in taxes.
Now back to Scenario 1 (Do *NOT* convert any additional): However, if I choose *not* to convert the extra $88,100, and just left it in the Trad IRA, it’d grow over the next 20 years, Tax-Deferred. If that $88,100 doubles in 20 years (that’s the conservative projection) the balance would be $176,200. However, I’m a pretty aggressive long-term investor, so let’s go with the historical average (doubling every seven years), which means that $88,100 would grow to about $528,600. My heirs would owe taxes on that.
*
What bracket might *they* be in? My sons are currently age 22 and 23. I believe they will be in *at least* the 32% bracket, probably higher. Here’s why: By the time I “keel over” in 20 years, if all goes reasonably well for them careerwise, they will be in their peak earning years. So let’s assume the 32% bracket married, filing jointly. But in addition to their work, they’ll inherit three buckets of assets: (1) Roth IRA (current balance $435k, which they must deplete within ten years). (2) Tax-Deferred IRA (current balance $2.2M, with mandatory RMDs); (3) Taxable Account current balance $2.1M – RMDs are not applicable). The first two of these could increase their tax brackets. Let’s assume these current balances will increase in 20 years, so the required Roth IRA withdrawals and the RMDs will be significant. There’s also the probability of higher tax brackets in the future (today’s National Debt is 26 trillion, up from 20 trillion just four months ago. Just sayin’).
From all this this I venture that they will be in *at least* the 32% bracket, which is *higher* than the 24% bracket am in today – and the *same* as the 32% I’m considering in Scenario 2.
On the other hand, WCI’s second point about them possibly stretching the Roth for 10 years is well taken: Yes, they might just let the Roth stretch, take the necessary RMD’s from the tax-Deferred account, and be in a position to retire early (or work part time) – in their early 50’s (gasp, my babies, how did that happen!). In that case it’s possible they could be in a bracket that’s relatively *lower* than the 32% we’re talking about. If that’s what they do – there’s no knowing, of course, that’s 20/30 years away – then Scenario 1 (*not* do the additional Conversion) makes more sense. So what to do?
** ** **
The other part of the equation is *my* tax bracket over these coming five years. I believe I’m in a “low tax bracket window” right now – and will be for the next 3 ½ to 5 years. Here’s why: To clarify Financial Dave’s comment, I am 66 ½, no longer working, (no possibility of Roth Contributions – only Conversions), and our income consists of my wife’s Government pension ($80k gross), plus rental property income ($40k gross), for a total of $120k gross.
Here are the time “windows”: Window 1, before starting to take Social Security income at age 70, just 3 ½ years from now (extra gross income of $40k/year, per the IRS). This is the *best* window, and gives me the opportunity to convert a lot to a Roth. Window 2: before my RMDs kick in at age 72, five years from now (approx. $80k/year). So, those two things will boost the income from today’s $120k gross, to $240k gross – double what it is today. At that point, I’ll still be inside the 24% of course, but the bad news is that the amount of Roth Conversion I can do, if I want to stay within 24% – will be decreased. So that’ll significantly slow down any Roth Conversion plan – unless of course (here we go again) I choose to bump up into the 32% bracket. It’ll be right around that time – the year 2025 – that the current tax code gets revisited, and taxes likely increase. So, all of that makes me say: better to bite the bullet, get ahead of all that, and bump up into 32% *now.*
*
To your other points, Dave: I said “in 20 the $2.2m could grow to $4.4” — this is extremely conservative. More realistically I’d expect to see a much higher growth rate than that the 3.526% you mention. To me, higher growth expectations would *significantly strengthen* the case for Scenario 2 (converting *above and beyond* the CPA’s $215k suggestion).
Yes, Dave, you’re absolutely right, my thinking has been to take the next six (6) years (before I hit 72) to convert as much of the $2.2M as possible – ideally all of it by 2025/2026. And yes, over these coming years it’ll still be growing, so in order to convert it *all* by then I’d definitely have to be aggressive about that. So yes, with this plan, as you say, the Traditional IRA balance will decrease significantly, thereby reducing my RMDs – that’s part of the case for doing this, and additional support for Scenario 2.
** ** **
Again, the goal is to leave a legacy that doesn’t get shredded mercilessly by taxes 20 years from now.
*
In sum: My tax bracket is lower now than it will ever be, and the windows are closing. My sons’ tax brackets will likely be higher than mine is now, even if I choose to go to 32%.
So, what do we conclude from all this? Best to stay within the 24% bracket per my CPA — or is there actually a strong case for proactively going one level up into the 32% bracket – while these two “low tax windows” are open? Is my thinking still flawed?
Thank you again. I appreciate any thoughts.
** ** ** ** ** ** **
Lots of assumptions there that could go bad on you, but if you really think your sons will be paying 32% on that inheritance, better for you to convert it now at any rate up to and including 32%.
Thank you.
Yes, there are definitely some assumptions here, and associated risk – no doubt.
It seems likely to me that my heirs *will be* in the 32% marginal bracket, so it’d make sense for me to bite the bullet and choose to move myself up from 24% to 32%, even though that is counterintuitive and flies in the face of conventional wisdom. But I’d even go further, and say that this is still advantageous if it turns out years from now that they’re in the 24% bracket. Here’s why:
Let’s say I decide to do the *additional* conversion (beyond the $215k my CPA said keeps me just within the 24% bracket (gross income up to $326,600, married, filing jointly). In this case, I’d be making a conscious choice to go into the upper end of the *32%* bracket (income up to $414,700, married, filing jointly). So the *additional* amount I’d be converting, beyond the original $215k, would be $88,100. The $32% tax due on that amount – which I’d pay now from a Taxable account – would be $28,192. So how might that $88,100 in the Roth grow over 20 years? If we go with the historical average (doubling every seven years), that additional $88,100 in the Roth would grow to about $528,600 – and it’d be *Tax-Free.*
However, if I choose *not* to convert that additional $88,100, and *just leave it in the Trad IRA* instead, it would similarly grow over the next 20 years – to $528,600 – but the bad news is it’d be *Taxable.*
So, how much is paid in tax, and what’s the net balance (legacy) when all is said and done?:
If they’re at the 32% rate, I’d have paid ZERO in taxes, and they’d pay $170k in taxes on that $528,600, so the total taxes we pay, between us, is $170k. And they end up with a *net of $360k.*
Even if I am wrong about their future tax brackets — and it turns out that they are in a bracket that’s *lower than* 32% (let’s say it’s 24%), it would seem that I have goofed. But, in fact, it’s not a problem — I’d have paid ZERO in taxes, and they’d pay only $125k in taxes on that $528,600, so the total taxes we pay, between us, is only $125k. And they end up with a *net of $400k.*
The point is this: whether they’re in the 32% or the 24% bracket,*the Roth still works out better* due to the fact that the Roth ending balance grows over 20 years (an assumption, for sure). The Roth is better for two reasons: (1) between us we pay less taxes with the Roth (I pay a little, and they pay ZERO), and (2) the Roth amount they net is higher — $528,600 *Tax-Free.*
** **
(But, yes, admittedly, because I’d be using the $28,192 to pay taxes *now* from my taxable account I do lose its growth potential. Assuming the same grow rates as above, that $28,192, had it been simply left in the Taxable account, would have grown in 20 years to $170k. However, it would be taxable).
What am I overlooking?
Again, thank you very much for your comments. They are greatly appreciated.
you forget that part of the doubling each 7 years (in most investments) is annual dividends etc, and instead of reinvesting tax free (Roth) you lose 10-20+% of that amount in taxes each year- the annual tax usually due from you (10- 20% qualified div rate x 8% of total or so, or 1+% drag) until they inherit.
But is that counteracted by the chance of your heirs stepping up the basis of the taxable account to zero- so you pay 1-2% taxes (tax on earnings ) then the IRS pays the capital gains, which are lower anyway?) I’ll do the specific math myself in the next few decade after retiring as my tax rate tops 24% only if I convert a whole lot more annually, and compare no more taxes but 10 year limit on payout to stepup in basis then taxes only on annual earnings until heirs wish to pay capital gains taxes (on $0 in the first year).
If I end up using it all, it is probably better to Roth it all so I avoid capital gains on up to 50 years of growth. If they spend it right away, then either one is good? No need to compare regular tax rate for heir to MY capital gain rate, correct? I should never pay more in capital gains (to get the cash to pay the tax to convert to Roth) than the Roth will save at least ME in taxes through my life expectancy? I think I need a believable excel sheet to play with the decision- but personally only if I build it myself will I be certain I comprehend all the variables.
Jenn,
” I should never pay more in capital gains (to get the cash to pay the tax to convert to Roth) than the Roth will save at least ME in taxes through my life expectancy? ”
If you pay $1 in capital gains tax to sell your funds to pay the tax, the taxable account loses and so does the conversion for the situation where we are assuming your ordinary income tax on the IRA withdrawal to do the conversion is the same as your tax rate in retirement – giving no advantage to IRA or Roth (tax rates equal). The time your money stays in the Roth has nothing to do with the outcome if ordinary income tax rates on the IRA withdrawal and retirement withdrawal later are the same.
It may help to first understand that TIME is not a factor in whether IRA or Roth wins.
https://seekingalpha.com/instablog/3752451-financialdave/4845086-roth-vs-non-roth-401k-403b-457-etc-time-value-of-money
After that what you need to know is having a taxable account is not going to win over having a Roth account unless once again there is some tax difference into the Roth vs out of traditional IRA in retirement. A taxable account just adds another layer of tax on top of what you were already paying. The money was already taxed going into the taxable account and in your case is now also being taxed for capital gains when you withdraw it.
It’s pretty easy to build a spreadsheet but you have to follow the money from what I like to call “cradle to grave.” Most people forget taxable accounts are very much like a Roth account until you have to pay tax of any kind, then they are not so good. The fact that you only pay a lower capital gains rate only means they are slightly “less bad.”
I wouldn’t go that far. The effects of tax drag are not insubstantial over decades of compounding. Even in a very tax efficient fund like TSM, it could be 0.5% a year. It would be higher with less tax efficient investments.
You’re overlooking the money used to pay the taxes on the Roth conversion. In your second scenario (no conversion) you’re not giving that to your kids. If you adjust for that, you’ll arrive at a different conclusion.
Derrick,
I think you are worrying too much about something that is totally outside your control – trying to optimize the money your heirs get and not thinking about the REAL risks in this situation. Don’t feel bad everyone tries to do the same thing when in reality there are two very important rules that come before everything else in this order of importance:
1. First, both halves of the married couple need to survive, hopefully with enough money to pay for “eldercare” or unforeseen medical problems or financial disasters that drain your account. That is job #1.
2. When it comes to your kids’ inheritance if they are very well off they won’t mind paying a little extra tax to get money they don’t need in the first place. To be perfectly honest most of my work suggests the difference of one or two tax brackets won’t make that much difference overall. Remember this decision is how to help them out the most. I would venture a thought that you can’t predict with any certainty how your kids will be doing 30 years from now, so why not plan to help them when they might need it most. If they are out of work and can’t find a job! In that case, I think you can figure out which path is best for you to do that — give them most of the money in an IRA rather than a much smaller amount in a Roth, since they will have a very low tax rate.
In fact, both #1 (keeping yourself solvent) and #2 (helping your kids when they need it) suggest not making the IRS rich by converting most of your IRA now is the best course of action. The side of the risks with (IMHO) all the bad outcomes originate because people are so worried about taxes later on that they send TOO MUCH money to the IRS on the front end and then have no options later on.
WCI,
<>
I think that is pretty much what I said – tax in a taxable account is bad and yes there are varying degrees. It is also possible to have NO tax in a taxable account for decades owning stocks like BRK.B, ILMN and others that could equal, or surpass TSM, step-up on death and pay no tax at all. It’s just a matter of degree.
Thank you again, Dave and WCI – excellent points. To respond:
Dave, for any emergencies and/or “eldercare” I’d use our Online Savings account, then the Taxable account, which together have a combined balance of $2.2m. If that’s not enough, God forbid, we’d also still have access to the Tax-Deferred IRA and also to the Roth. And also Social Security and RMDs when they start, three and five years from now. Your point is well taken, but I believe there’s a decent cushion.
Yes, Dave, I totally agree there’s no knowing how my kids will be doing 30 years from now, nor do we have any clue what their future tax rates will be. I’d like to think that we *are* currently in a position to help them out over these closer-in years if/when they need it (as you suggest, perhaps in times of unemployment; or to provide an AFR-appropriate loan for down payment on a house purchase, etc.), as well as through some gifting that’s below/in line with the $15k limit, to help them build some investing skills and to foster their personal responsibility. So yes, my plan is not only to leave them a tax-efficient legacy, *but also* to help them in these new, post-college years, as appropriate.
**
Switching gears …
The discussion so far has focused on my bracket *today* and what I imagine my sons’ distant future brackets will be whenever they inherit. I suspect that my current 24% rate is lower than theirs will be, but yes, your point is well-taken: their future tax brackets can only be a guess.
Setting that future-oriented discussion aside, I have also started to look closer in — specifically at my bracket *today* and comparing it to what it might be in the next 3 – 5 years, (taking into account my upcoming Social Security and RMDs). There’s a strong likelihood that today’s current, historically low brackets will expire the end of 2025, whereupon we revert to the pre-2018 (higher) levels. We currently have an out-of-control National Debt, trillions in new spending to alleviate the market turmoil brought upon us by Covid-19, and other financial messes. So, question: The current low tax rates have been (and are) a gift, but would you agree that this gift will *not* be extended beyond 2026?
That’s relevant because I’m in the 24% bracket, and with today’s gross income of $130k I am well within the top end of the 24% bracket (up to $326,600). That’s today, but, even when I start to take Social Security and RMDs, *and* even when I do a sizeable Roth Conversion, I’m *still* inside 24%.
However, I’m going to assume that on January 1st, 2026, the brackets will revert to the pre-2018 levels, and that we’ll suddenly be at 33%. This is because (1) the tax rates change, but (2) also because my wife starts *her* Social Security in 2026. So those two factors will put us over the top.
This “low-tax window” (from now until Congress changes the tax brackets in January 2026) is obviously a great time to do a series of Roth Conversions – up to 24% as already discussed. The question remains, though: whether to convert an *additional* $88,100 in 2020, thereby voluntarily bumping up into marginal 32%.
My latest thinking – after running some detailed numbers on income from my forthcoming Social Security, and my forthcoming RMDs, etc.), I only see *a very minor upside advantage* to converting the additional $88,100 in 2020, and voluntarily bumping up into the 32%. Here’s why: Before 2025 we’ll be in 24%. If Congress *does* change the brackets in 2025 (likely), we’ll be firmly in the New 33%. The New (post-2025) 33% bracket is only 1% higher (i.e., a bit more tax) than the 32% I would convert at today — so there’s a tiny upside of 1% tax savings.
On the other hand, on the remote off-chance that Congress does *not* change the brackets, we’ll remain within 24% for the foreseeable future – even *after* 2025. In that case it’d be foolish to pay 32% now on that additional $88,100: there’s greater downside risk (8% overpayment: 32% versus 24%) than upside potential (1%: 32% versus the new 33%). *So, in summary, the decision, it seems, comes down to whether or not the rates will revert back at the end of 2025*. Nobody knows, of course, but I venture they *will.*
So, question: Other than that potential tiny 1% upside, are there any *other* reasons to consider doing the *additional* $88,100 conversion? E.g., reducing the overall size of the Taxable Estate, by paying Uncle Sam now, with the assumption that over a decade or two the Estate balance will grow such that it may well exceed the Estate Tax threshold (currently $11.58 million per person, but also likely to be repealed to pre-2018 levels)?
A good problem to have, I know, and I am grateful. Thank you again for your comments.
Derrick,
“The point is this: whether they’re in the 32% or the 24% bracket,*the Roth still works out better*
Not really, it doesn’t really matter who pays the tax, you have to think of the process as “money you earned” flows towards “somebody spending it.” How much after-tax money there is to spend (more or less) relates mostly to the tax difference. TIME is not a factor for how long the money stays in the Roth until you figure out whether that “time” is compounding your spendable income UP or DOWN.
Also, remember that in retirement you spend money across multiple tax brackets starting at whatever your base (or your kids) is. If your base is only SS (and not pensions) then your base essentially starts at a zero tax rate but goes up faster until 85% of your SS is taxed and then it slows back down.
Personally, I don’t know where taxes would go, but I would put less than 50% chance that they get rolled back to 2018. You also forget that tax-brackets are inflated by about 2% per year so by 2026 the top of the 22%, 24% etc, bracket will be higher.
I see nothing wrong with converting some money, but I don’t see a case to be made that you need much more than 25% tax-free vs tax-deferred and I absolutely believe converting it even at 24% is reducing the after-tax spending income for most reasonable cases of where the money is going to be spent — depending on what State taxes you have, as we only seem to be talking about Federal.
PS. I rarely try to make too many decisions based on future unknowns. It’s best to just “diversify” your decision in cases like this.
Thanks for this post! Do you know if the changes are retroactive for an IRA that was inherited years ago or just to newly inherited IRAs moving forward?
Just Googled it- only applies to accounts inherited from someone who died after December 31, 2019.
Older inherited IRAs are grandfathered in to the old rules.
If one is concerned about heirs blowing inherited retirement money, then the solution is to make a trust the beneficiary of those assets. One can limit distributions from the trust. The money would still be forced out of the retirement account but the trustee or the terms of the trust could keep it from being rapidly depleted.
As it turns out I studied a few of the results of the new stretch IRA after the Secure act. While the facts support that only the lucky few will ever receive over $1m, I studied $1m and $2m portfolios and it turns out that optimizing for the maximum after-tax amount is very difficult because it depends on knowing in advance what the returns are going to be over the 10 years.
A couple of the better strategies are:
With a 3-4% return start in year 5 and withdraw 5%, 10%, 20%, 30%, 50%, 100%
With no growth your best bet is just 20% for 9 years and 100% the last year to close it out.
Interesting thought experiment, but it definitely relies heavily on whatever assumptions you make. For instance, if you are in the top bracket already, the best idea is to leave it all there until year 10.
I agree, but if you’re making $630k per year maybe your best bet is to disclaim it and let it go to another relative if there is one.
I received my portion of a small traditional 401k (about 17k) last year after my father passed away suddenly. My spouse will be finishing up his last fellowship year July 2021. We were planning on doing our Roth conversions in his last year of training anyway before attending income, but especially now with the market drops we are thinking we can take advantage of taxes being “on sale.” We have been working on doing Roth conversions of all our traditional IRA/401k accounts (about 52k including the inherited 401k). The blog post mentioned just leaving a traditional IRA where it is if it is small, but this particular employer’s 401k plan would require I transfer out the money in 5 years. I was planning on rolling it into a traditional IRA, then doing the Roth conversion (I’ve been told it’s better to do it in 2 steps.) We have 3 kids and I don’t work so we have a very, very low tax rate as my husband will be on a training salary only through 2020. Even though it’s a small account, I would think it would make the most sense to convert to Roth now still as our tax rate will be substantially higher in 10 years (would be true of any attending salary but he is in a higher paying specialty). We are in our mid-thirties and I would never consider cashing out the policy since I wouldn’t want to pay the early withdrawal penalties and I want the account to grow tax-free while it can. Are we on the right track on this?
You can’t do Roth conversions on inherited 401(k)/IRAs. Sorry. You should be able to roll it over into an inherited IRA though and stretch it out a bit longer before withdrawing, but given your current tax rate, it may be smarter to just take it all out now and reinvest it in your own investing accounts.
There are no early withdrawal penalties on an inherited IRA either.
Thank you. I had not thought through the benefits of cashing out and re-investing but this does seem like the best option for us.
My condolences on the loss of your father. I understand your concerns, as I have managed similar situations. If you are not maxing out your spouse’s 401k/403b/457b retirement plan and/or both of your ROTH IRAs, you could withdraw the approximate $17K from the inherited 401k, pay the necessary taxes, then increase the contributions to your spouse’s employer’s plans and/or both of your respective existing or newly created ROTH IRAs. If your spouse’s employer does not offer a ROTH option, I think your best bet is to contribute $6K to your spouse’s ROTH and $6K to your spousal ROTH. Where you put the remaining $5K would depend on your overall financial picture, with it either increasing contributions to your spouse’s deferred plan or to a taxable investment account. This approach would generate a rough equivalent of a conversion, if conversions were allowed. If you already max everything out, then investing the funds in a taxable account is likely your best option. Hope this helps.
This makes total sense- I really appreciate your advice. I plan to cash out the 401k and invest in IRAs for both of us for 2020 and will think about where else to invest the remainder.
There are some significant errors in this post. First of all, Roth IRAs do not currently have RMDs or taxation on distributions, so the paragraph with this heading mixes apples and oranges:
Change # 1 No RMDs Required for Inherited IRAs. The first change, which everyone seems to gloss over, is the elimination of a requirement to take an RMD from the inherited IRA every year. You don’t have to take a single withdrawal from that sucker until year 10. That is a substantial advantage. Consider someone who leaves their Roth IRA to their 80-year-old little sister.
Inherited IRAs and Inherited Roth IRAs are two different beasts from a taxation point of view. And most are not left to a younger sibling, but to a child.
Also, non-spouse beneficiaries are not “Eligible Designated Beneficiaries” and this article makes it seem like you can make them so. You can’t. After reading this, this seemed more like “clickbait” rather than actual helpful info.
What are you talking about? Prior to the change, inherited Roth IRAs absolutely did have RMDs. The headline is accurate as it stands and is hardly an error.
The paragraph also does not state that most IRAs are left to a younger sibling. It says, “Consider someone who leaves their Roth IRA to their…sister.” That’s not an error.
As far as “eligible designated beneficiaries”:
“Any of the following individuals are considered an eligible designated beneficiary (EDB): a surviving spouse, a disabled or chronically ill individual, an individual who is not more than 10 years younger than the IRA owner, or a child of the IRA owner who has not reached the age of majority.”
http://www.wolterskluwerfs.com/article/secure-act–changes-distribution-requirements-for-some-individuals.aspx#:~:text=Any%20of%20the%20following%20individuals,reached%20the%20age%20of%20majority.
https://www.kitces.com/blog/secure-act-stretch-ira-401k-elimination-eligible-designated-beneficiary-retirement-accounts-taxes/
As you can see, there are beneficiaries who are not spouses who ARE eligible designated beneficiaries. Thus, this is also not an error.
Finally, clickbait is not an insult around here. It is evidence of a title well written. But if you don’t think my articles are helpful, stop reading them, much less leaving comments on them saying they contain errors when they do not.
Please clarify at the beginning of the blog that this only applies to IRAs inherited after Dec. 31, 2019. I panicked thinking I was going to have to withdraw everything from mine in 4 years until I looked it up and saw I was grandfathered (should have assumed so, but would be nice to state that up front).
Has there been changes to this again, vanguard seems to indicate currently that whether roth or traditional, spousal or non-spousal – you are required to take RMDs.
No changes. No RMD required for 10 years on inherited IRAs. Spouses never have or had to take RMDs from inherited IRAs. Got a link to the Vanguard piece and maybe I can help point out the issue with it.
What do you think about this strategy that Ed Slott among others touts wherein you –
1) Change the beneficiary of your IRA to a Charity
2) Withdraw enough money from the IRA every year to pay taxes on the withdrawl and use the rest to pay a premium on a Universal 2nd to die Life Insurance policy with a death benefit equal to what you think the IRA will be worh when the 2nd dies.
This allows you to give to Charity and to your heirs tax free…
It seems shady but I can’t figure out why I don’t like it.
Nothing shady about it. Perfectly legal. I would argue it isn’t very smart, but there’s nothing illegal about it. That life insurance policy is likely to grow slower than if you just invested the money if you die at your life expectancy. If you die before then your heirs could come out ahead though.
And of course tax-deferred money is best left to charity if you have charitable inclinations. But anything you leave to charity is tax free to the charity. They don’t pay taxes.