Monster Roth Conversions - How big to go and for how long?

[Dash man]

We’re in a similar situation to you. We’re also considering that today’s tax rates will end after 2025, reverting to the previous tax rates in 2026. So we are converting well into the 33% right now.
We’re also looking at the possibility of a survivor having to pay taxes at the single rate. Another reason to move more sooner.
There is also the possibility of Federal Estate taxes and capital gains taxes going up. While this will require changes in legislation and may never happen, the two combined would be a major hit on your heirs, which can be lessened by using your taxable funds to pay the taxes on more conversions.

 

[Montecfo]

If you are trying to reduce overall taxes I think in this case some consideration of heirs' likely tax rates is appropriate.

Not sure I would voluntarily pay taxes in the higher brackets since you are not going to reduce RMDs much and are already in high brackets. But the surviving spouse issue is a large consideration.

Quite a few moving parts here.

 

[RunningBum]

OP has mentioned having "run the math". What are the actual numbers from that?

From what I understand, you'll have 14 years to do conversions before RMDs. I'm guessing you have room to convert ~$300K to the top of the 24% bracket. That depends on how much other regular income you have, and what deductions you can take. It may be more than $300K.

14 years of that means you can convert $4.2M. That's a pretty darned good chunk! The tIRA will be growing as you convert. Does your math include a spreadsheet showing growth of the tIRA while you convert? You could model a few different assumptions on growth.

Once you've run this it should be pretty easy to compare it to doing no conversions, and seeing how large the tIRA grows that way. It will be quite a bit more. People seem to get hung up with not really making progress on converting a large tIRA, but if you compare it to doing no conversions you'll see what progress you really are making.

Then look at what RMDs will be in each situation. This will tell you whether it is worth converting into the 32% bracket. Without seeing assumptions and numbers we're really just guessing on the 32% question. If it's close, I'd be inclined to go to the top of 32% for a few years with the threat of higher taxes resuming in 2026, and dropping back if taxes do go higher.

 

[Gumby]

Regarding just taking tIRA distributions versus doing a Roth conversion, I look at my Roth account as a conveyer belt. That is, I pour new conversion money in one end and take tax free money (which has hopefully appreciated) out the other end. I can run the Roth conveyor belt as fast or slow as I like, but the whole time the money is on that belt it is earning on a tax free basis. If I were just to take a tIRA distribution, I don't get that benefit.

 

[sengsational]

Running i-orp is always what I recommend in these cases because it models all years in one swell foop. As with any model, it's not going to match reality perfectly, but if you take the time to get the inputs defined properly, it might give you an eye-opening approach that you hadn't thought about. You can then try that approach in your own spreadsheet model and see if you like it or not. As mentioned, the bird in the hand problem (definitely save taxes now vs maybe save taxes later) tends to become a problem when, with the best assumptions, your early conversions are frightfully high.

 

[SecondCor521]

OP has mentioned having "run the math". What are the actual numbers from that?

From what I understand, you'll have 14 years to do conversions before RMDs. I'm guessing you have room to convert ~$300K to the top of the 24% bracket. That depends on how much other regular income you have, and what deductions you can take. It may be more than $300K.

14 years of that means you can convert $4.2M. That's a pretty darned good chunk! The tIRA will be growing as you convert. Does your math include a spreadsheet showing growth of the tIRA while you convert? You could model a few different assumptions on growth.

Once you've run this it should be pretty easy to compare it to doing no conversions, and seeing how large the tIRA grows that way. It will be quite a bit more. People seem to get hung up with not really making progress on converting a large tIRA, but if you compare it to doing no conversions you'll see what progress you really are making.

Then look at what RMDs will be in each situation. This will tell you whether it is worth converting into the 32% bracket. Without seeing assumptions and numbers we're really just guessing on the 32% question. If it's close, I'd be inclined to go to the top of 32% for a few years with the threat of higher taxes resuming in 2026, and dropping back if taxes do go higher.

I agree with the first part of your post.

Regarding the last paragraph of your post, it's pretty easy to do a rough swag using nominal numbers and the rule of 70/72:

OP appears to have a $6M 401(k). Fourteen/fifteen years and the rule of 70/72 assuming 10% return (people can argue with that assumption, OP should substitute their own opinion) implies two doublings, so that becomes $24M at RMD time.

The new age 72 divisor is 27.4. $24M / 27.4 = $876K. The top of the 35% bracket in 2021 is $523,600. Brackets are adjusted for inflation; at 2% per year (again, OP, make your own assumption) for 15 years that works out to about $705K in 2036.

OP's and his wife's SS will probably more than wipe out OP's standard deduction.

IRMAA makes the situation even worse.

State taxes are probably a wash unless OP plans to move to another state.

The above is a rough estimate, but if I were in OP's shoes, I'd convert to the top of the 32% bracket, especially if I could pay the taxes from the taxable account (which OP looks like they can do).

Logically it might even make sense to convert some into the 35% bracket, although the savings there are probably minimal and harder to make an argument for.

 

[SecondCor521]

OP, another thing that I'll point out in a separate post, which someone else alluded to above.

In addition to the income tax torpedo you're facing, you very well might also be facing an estate tax torpedo.

Today, the estate tax exemption amount is about twice your investable assets. For estate taxes, you'd also add in the value of any homes, cars, or life insurance face value that is owned in your names.

So again, using hand-wavy math and my assumptions from the previous post, your taxable estate will probably pass the exemption amount. Every dollar past that amount costs you 40% in estate taxes.

If you have heirs other than charity, they're losing out on that 40% in taxes.

It gets worse. The current historically large exemption amount is scheduled to drop approximately in half on 1/1/2026. At that point, you'll have about (very hand-wavy) $20M and the exemption amount will be about $7M. $13M times 40% is a $5.2M estate tax bill.

It may get even worse. There is a policy proposal to reduce it sooner to $3.5M and raise the estate tax rate to 45%. If that were to be enacted into law, you've got more like a $6.5M estate tax bill.

Some states pile on with estate taxes and inheritance taxes of their own.

I won't go beyond this into all the ins and outs of estate tax that I've been learning about. I do have some things I'm working on for my own family's situation to help out, and I can share those with you if you want. I'd just close by saying that it's something that may loom large in your future and starting to steer your financial ship now to avoid it is probably a wise option to consider.

Good luck.

 

[DawgMan]

You guys are killing me!

And I thought it was a good thing to make/save money!

 

[RunningBum]

You keep shooting messengers, you won't have any left.

Between that and going back and forth on trying to be tax efficient and not worrying about it because you have enough no matter what has me wondering why I'm bothering here.

 

[fiyesno]

Here is something I have been trying to work out related to this thread. Let’s say you have a goal to give away a million dollars while you are still alive. How about moving post-tax account appreciated assets into a donor-advised fund (Fidelity and Schwab both have easy ones to use) at the same time as doing large roth conversions. For example, convert $500k 401k to roth in a year, and move $250k taxable (with say a $50k cost basis) to a DAF. You would then have a taxable net of $250k from these conversations. It seems like this might work well and allow you to convert a large amount to roth in a small number of years. But I am not totally sure how to model this out accurately.

Has anyone modeled this type of scenario out?

 

[RunningBum]

Here is something I have been trying to work out related to this thread. Let’s say you have a goal to give away a million dollars while you are still alive. How about moving post-tax account appreciated assets into a donor-advised fund (Fidelity and Schwab both have easy ones to use) at the same time as doing large roth conversions. For example, convert $500k 401k to roth in a year, and move $250k taxable (with say a $50k cost basis) to a DAF. You would then have a taxable net of $250k from these conversations. It seems like this might work well and allow you to convert a large amount to roth in a small number of years. But I am not totally sure how to model this out accurately.

Has anyone modeled this type of scenario out?

This seems barely related to this thread and would be better in a new thread, IMO. Search for DAF in this forum and you might get some answers. If you still have questions, I'll give my experience in a new thread. I got this idea from others on here so you might get other answers. Very short answer, not sure what you are looking for on modeling but it seems like the best way is to do it is to get the largest tax break, so a DAF (with appreciated shares as you say) or QCDs are most likely the top two methods, unless you get into something more complicated like a charitable remainder trust.

 

[DawgMan]

You keep shooting messengers, you won't have any left.

Between that and going back and forth on trying to be tax efficient and not worrying about it because you have enough no matter what has me wondering why I'm bothering here.

Easy does it, only joking around. Just trying to get my arms around all the moving parts. Your suggestions are appreciated.

 

[SecondCor521]

You guys are killing me!

In the good news department, you'll avoid estate tax issues that way.

 

[DawgMan]

To get a feel for the answser, you might consider running https://www.i-orp.com/Plans/extended.html . Make sure to check off the option of "IRA to Roth Conversions = 7 Unlimited". The program will suggest optimum Roth conversions if you spent the maximum allowable amount that your savings allow for the life time you input.

I took a shot with my interpretation of your numbers. I-ORP suggests you could spend of~ $590,000 per year (adjusted up each year for cost of living) excluding taxes for the rest of your life. If you actually spent that much yearly and wanted to optimize lifetime conversions, ORP suggests convering into the 37% tax bracket for a couple years, and the 35/33% tax brackets for another 6-7 years.

From your comments, I expect you won't spend 590k / yr so optimum Roth conversions may be higher. Also I expect (IMHO) that tax rates in the future will be higher than the current tax laws so higher Roth conversions today may be beneficial. Note: I-ORP takes into account the current tax rates and the changes when they expire in 2025.

Many have limits on how much tax they are willing to pay today and don't convert as much as I-ORP suggests. I personally have come around to accepting I have to pay sooner or later and might as well try to optimize it as much as I can.

You have a great problem to have, congrats.

I took a shot at IORP as suggested and perhaps I am entering something wrong/differently (probably by more conservative annual returns), but I am getting the following output relative to Roth conversion years...

Age RMD TaxDef AfterTax RothIRA IRA2Roth GuarInc Yield Taxes DI
057 0 516 565 0 516 0 130 224 471
058 0 494 572 0 494 0 121 212 480
059 0 345 525 0 345 0 112 147 490
060 0 364 550 0 364 0 102 152 499
061 0 384 576 0 384 0 91 158 509
062 0 414 628 0 414 0 79 187 520
063 0 225 570 0 225 0 67 107 530
064 0 243 597 0 243 0 55 112 541
065 0 265 636 0 265 0 41 126 551
066 0 286 667 0 286 0 27 131 562
067 0 308 700 0 308 0 11 137 574
068 0 325 458 0 76 0 0 122 585

I'm somewhat new to IORP so may be missing something. Yr1 it's showing $471K in after tax distributions. Can I manipulate the model to only show say $250K in after tax distributions to see what it does to my Roth conversions modeling?

 

[davebarnes]

I read this thread this morning.
I had my session with our tax prep guy at 1100. I got to experience the NIIT.
We talked about Roth conversions.
As "the poors" with only $1.4M in tIRAs, our life is a bit simpler.
I believe "America is great" because we seem to have the most complicated tax code in the universe.

 

[Dash man]

Here is something I have been trying to work out related to this thread. Let’s say you have a goal to give away a million dollars while you are still alive. How about moving post-tax account appreciated assets into a donor-advised fund (Fidelity and Schwab both have easy ones to use) at the same time as doing large roth conversions. For example, convert $500k 401k to roth in a year, and move $250k taxable (with say a $50k cost basis) to a DAF. You would then have a taxable net of $250k from these conversations. It seems like this might work well and allow you to convert a large amount to roth in a small number of years. But I am not totally sure how to model this out accurately.



Has anyone modeled this type of scenario out?

Just be aware you can only deduct donations to a DAF up to 30% of your adjusted gross income.

 

[Out-to-Lunch]

Let's make sure we are using the same terms. In year 1, it is showing, I believe, $565k in distributions from your taxable account. It has your after-tax disposable income (DI) as $471. Is this the way you are understanding it, too?

The premise of I-ORP is to see how much DI you can have, and have it the same every year. It does not expect you to want to spend less than you can.

A tweak that you can do: You can specify that you will die with a lot of money (by specifying a large Plan Surplus).

 

[DawgMan]

Let's make sure we are using the same terms. In year 1, it is showing, I believe, $565k in distributions from your taxable account. It has your after-tax disposable income (DI) as $471. Is this the way you are understanding it, too?

The premise of I-ORP is to see how much DI you can have, and have it the same every year. It does not expect you to want to spend less than you can.

A tweak that you can do: You can specify that you will die with a lot of money (by specifying a large Plan Surplus).

You are correct. Wrong terminology. If I play around with the surplus to hit my desired DI, does it still model the most efficient Roth conversions (assumes I want to leave Roth/after tax accounts first)?

 

[Out-to-Lunch]

You are correct. Wrong terminology. If I play around with the surplus to hit my desired DI, does it still model the most efficient Roth conversions (assumes I want to leave Roth/after tax accounts first)?

That is a good question; I realized that not only did I not know the answer, I didn't have clarity on what I-Orp was trying to do. So I poked around, and found this helpful bit:

(You can find the explanation of the entry fields by clicking on the blue link.)

Minimum Roth IRA Balance

The Roth IRA minimum balance parameter is used to instruct ORP to maintain a specified balance in the combined Roth IRA accounts throughout retirement.

In Which Assets to Leave to Heirs and Related Issues Potts and Reichenstein [2015] demonstrate situations where there are tax advantages to leaving the estate all or partly in the Roth IRA Account.

Leave money in your Tax-deferred Account for heirs in income tax brackets lower than yours. Their tax-deferred withdrawals pay less income tax than you would. You can only do this to a limited extent because the Required Minimum Distribution (RMD) will drive your Tax-deferred account balance toward zero at your planning horizon.
Leave money in the Tax-deferred Account if you have a designated beneficiary who can withdraw the money at their RMD rate. This stretches withdrawals over anywhere from 44 to 66 years.
Leave money in the Roth IRA for heirs in income tax brackets higher than yours. There are no taxes on your heirs' withdrawals.
Money in the After-tax Account has it's yields subject to personal incomer tax each year. Txable account capital is not taxed until it is withdrawn when it is subject to capital gains tax.
For the details see Vanguard's discussion on the topic.

Left to its own devices ORP will leave all of the specified estate in the Tax-deferred Account because if the balance is not withdrawn there is no income tax paid -- by you. Your heirs, however, will have to pay income tax on their distributions from your Tax-deferred account.

ORP will maintain the specified minimum in your Roth for each year of retirement so that if you and your spouse do not actually live to the end of your plan, the specified minimum will be in the estate. The specified Roth minimum balance is inflated over the term of the plan.

The Tax-deferred Account does not have a minimum balance. ORP will leave as much of your Final Total Asset Balance (FTAB) as it can in the Tax-deferred Account. The RMD will drive the Tax-deferred Account balance toward zero over the course of retirement. This may conflict with a Roth IRA minimum account balance.

ORP will position at least your specified Roth minimum balance in your Roth and divide the remaining FTAB across all three accounts, depending on IRA to Roth conversions, compliance with the RMD and transfers to and savings in the After-tax Account.

Careless use of the Roth IRA minimum balance has the potential to create model inconsistencies that result in unsolvable models. When ORP declares your model to be unsolvable your Roth IRA minimum balance and your estate are the first places to look for the cause.

When you specify a Roth IRA minimum balance that is greater than the starting Roth account balances then ORP automatically enables unlimited IRA to Roth IRA conversions to ensure sufficient Roth IRA funds are available to satisfy your minimum.

 

[RetireBy90]

I read this thread this morning.
I had my session with our tax prep guy at 1100. I got to experience the NIIT.
We talked about Roth conversions.
As "the poors" with only $1.4M in tIRAs, our life is a bit simpler.
I believe "America is great" because we seem to have the most complicated tax code in the universe.

[RANT]As a common man with plenty to meet all our needs, I’m still surprised by the complexities of the federal tax system. I printed the 2020 tax last weekend, so I could compare key figures with the 2019 return, 69 pages ! Form 1040SR is up to 3 pages ! I couldn’t believe the info you have to provide that didn’t directly impact taxes ! [/RANT]

 

[Whisper66]

I took a shot at IORP as suggested and perhaps I am entering something wrong/differently (probably by more conservative annual returns), but I am getting the following output relative to Roth conversion years...

Age RMD TaxDef AfterTax RothIRA IRA2Roth GuarInc Yield Taxes DI
057 0 516 565 0 516 0 130 224 471
058 0 494 572 0 494 0 121 212 480
059 0 345 525 0 345 0 112 147 490
060 0 364 550 0 364 0 102 152 499
061 0 384 576 0 384 0 91 158 509
062 0 414 628 0 414 0 79 187 520
063 0 225 570 0 225 0 67 107 530
064 0 243 597 0 243 0 55 112 541
065 0 265 636 0 265 0 41 126 551
066 0 286 667 0 286 0 27 131 562
067 0 308 700 0 308 0 11 137 574
068 0 325 458 0 76 0 0 122 585

I'm somewhat new to IORP so may be missing something. Yr1 it's showing $471K in after tax distributions. Can I manipulate the model to only show say $250K in after tax distributions to see what it does to my Roth conversions modeling?

Sorry, been a busy day and just now catching up.

Your answers are in the ballpark of what I expected. Your results will differ from mine due to things I didn't know (such as SS amount, expected planning horizon, etc) and differing assumptions (example - I accepted defaults for rates of return).

The $471k value is what ORP calculates as your maximum allowable disposable annual income (excludes taxes). In other words, that's what you would need to spend spend annually to use up your assets fully over the planning horizon (estimated life time).

Your run suggests very high Roth conversions as did mine. See IRA2Roth column ... conversions starting at 516k, then 494k, etc... If you look at the last table in the report,, the Federal Tax Bracket report, you can see what tax brackets your Roth conversions are driving you into. I assume you will see it's well into the 30's brackets somewhere.

You only expect to spend about $250k annually. Directionally this will leave you with more money each year than the current run suggests so optimum Roth conversions will be higher. Some suggest that you just put some value in the "Plan Surplus" input to drive down the result for max disposable income. I'm not sure the progam can handle things well this way and have seen some funny results when I've tried it. So I've discounted that as an acceptable method for now.

I am retired and we don't spend near our "max disposable income" each year. So each new year, I run a new ORP case to see what it suggests the Roth conversion should be in the current year given I didn't spend as much as the prior year plan suggested. As expected, the amount of conversion ORP suggests for the current year is higher than the prior year results suggested I would need to make for the current year. This makes sense since I have more assets then the ORP run of the prior year figured I would have.

You could use the above concept to address your under spending the max disposable income. You could run multiple ORP cases. The first case would represent year 1, the next year 2, the next year 3, etc.... ( I think running 3-4 cases would give you a pretty solid understanding of what is happening with Roth conversion suggestions.) Run year one as you have already done but only consider the first years results. Then start a new case 2 to represent year 2. For input, use your best estimate for assets assuming your lower spend rate. And only look at the results for the 1st year in Case 2. Then do this again for case 3 representing year 3. Kind of a clunky work around but it's one way to manage it.

Another way may be to find a tool that allows your expected spend as an input. I haven't used it but hear others say there is another tool one can purchase that can do it. I think it may be called "Retirement Income Management Optimizer" ?

 

[Whisper66]

.....................You could run multiple ORP cases. The first case would represent year 1, the next year 2, the next year 3, etc....

Just thinking about this a bit ....

ORP suggests spending 471k in year one. You only spend $250k. This means you will have higher end of year assets than ORP expected by 471-250 = 221k (plus some earnings). That's 221k represents about 1.8% of 12M total assets. I expect you'll just see somewhat higher conversions suggested for somewhat more years. But only running the numbers will tell for sure.

 

[kjpliny]

With that much money, I'd be tempted to convert it all at 37% in one shot and be done. Then you can live the rest of your life giving the finger to the man. Seeing how much debt we're going into as a nation, you could find yourself getting taxed well over 50% in the future.

 

[davebarnes]

With that much money, I'd be tempted to convert it all at 37% in one shot and be done... [vs] getting taxed well over 50% in the future.

I think we will get a year+ warning about the 50% rate, so I plan to proceed at 24% until/if the hammer drops.

 

[SnowballCamper]

I am retired and we don't spend near our "max disposable income" each year. So each new year, I run a new ORP case to see what it suggests the Roth conversion should be in the current year given I didn't spend as much as the prior year plan suggested. As expected, the amount of conversion ORP suggests for the current year is higher than the prior year results suggested I would need to make for the current year. This makes sense since I have more assets then the ORP run of the prior year figured I would have.

You could use the above concept to address your under spending the max disposable income. You could run multiple ORP cases. The first case would represent year 1, the next year 2, the next year 3, etc.... ( I think running 3-4 cases would give you a pretty solid understanding of what is happening with Roth conversion suggestions.) Run year one as you have already done but only consider the first years results. Then start a new case 2 to represent year 2. For input, use your best estimate for assets assuming your lower spend rate. And only look at the results for the 1st year in Case 2. Then do this again for case 3 representing year 3. Kind of a clunky work around but it's one way to manage it.

When you run ORP you are telling the computer model to determine your maximum disposable income. It gives the best case scenario results, and this is helpful to determine IRA conversion strategies because higher income is taxed at a greater percentage.

The second part we should use is the Montecarlo simulation to see what the likely disposable income will be. ORP doesn't suggest spending the maximum...

I recall most of the posts on the forum from people who use ORP do an annual run to guide how high a tax bracket to to convert up to. I think that is prudent, but with a large increasing traditional IRA balance you could be doing conversions for a while. I think you can model a specified large conversion and compare the result to smaller conversions over more years. In each case, the prudent comparison is the Montecarlo simulation output, not just the maximum disposable income (which will show an inflated difference).

For my own case, there is little difference between doing conversions or not, and I'm not in a higher tax bracket, so not planning to convert.