Pre-tax vs Roth for HNW

Have a question about your personal investments? No matter how simple or complex, you can ask it here.
Luckywon
Posts: 2422
Joined: Tue Mar 28, 2017 10:33 am

Re: Pre-tax vs Roth for HNW

Post by Luckywon »

toddthebod wrote: Sun Jun 09, 2024 1:15 pm That is not how the IRD deduction is calculated. Based on my rudimentary understanding, I think it works as follows: you calculate the estate tax twice, once including the IRA, once without, and the difference is the deductible amount. In this case, since $15M - $3M is less than the annual exclusion, the entire $832,000 should be deductible.
Ahh that sheds a lot of light. I knew I must be missing something as bsteiner is always correct!! :beer
toddthebod
Posts: 6718
Joined: Wed May 18, 2022 12:42 pm

Re: Pre-tax vs Roth for HNW

Post by toddthebod »

Luckywon wrote: Sun Jun 09, 2024 2:05 pm
toddthebod wrote: Sun Jun 09, 2024 1:15 pm That is not how the IRD deduction is calculated. Based on my rudimentary understanding, I think it works as follows: you calculate the estate tax twice, once including the IRA, once without, and the difference is the deductible amount. In this case, since $15M - $3M is less than the annual exclusion, the entire $832,000 should be deductible.
Ahh that sheds a lot of light. I knew I must be missing something as bsteiner is always correct!! :beer
It looks way more complicated in reality, but I'm pretty sure the basic idea is right, that the deduction for the estate tax allocated to the IRA is based on the estate tax calculation with and without that IRA.
privateer79
Posts: 484
Joined: Fri Apr 04, 2008 12:21 am

Re: Pre-tax vs Roth for HNW

Post by privateer79 »

smitcat wrote: Sat Jun 08, 2024 11:19 am
privateer79 wrote: Sat Jun 08, 2024 11:05 am I wonder how much of this hangs on state tax rate arbitrage?

being in a ~15% state and local bracket + high federal, pretax lets me avoid the 15% state haircut by withdrawing later in a retiree-friendly state (florida, Arizona, etc)

being in a 0 % income tax state takes alot of that benefit away.

I was unaware that any state levied a 15% tax.
Geoarbitration works fairly well as long as you have a desire for those destinations.
Best to consider all of the costs when comparing any two locations not just the income taxes.
state + region + city... when you keep adding 1% here, 1% there, it starts adding up.
Luckywon
Posts: 2422
Joined: Tue Mar 28, 2017 10:33 am

Re: Pre-tax vs Roth for HNW

Post by Luckywon »

toddthebod wrote: Sun Jun 09, 2024 2:08 pm
Luckywon wrote: Sun Jun 09, 2024 2:05 pm
toddthebod wrote: Sun Jun 09, 2024 1:15 pm That is not how the IRD deduction is calculated. Based on my rudimentary understanding, I think it works as follows: you calculate the estate tax twice, once including the IRA, once without, and the difference is the deductible amount. In this case, since $15M - $3M is less than the annual exclusion, the entire $832,000 should be deductible.
Ahh that sheds a lot of light. I knew I must be missing something as bsteiner is always correct!! :beer
It looks way more complicated in reality, but I'm pretty sure the basic idea is right, that the deduction for the estate tax allocated to the IRA is based on the estate tax calculation with and without that IRA.
It seems then a way of looking at this is that tax deferred accounts are effectively not subject to federal estate tax?
toddthebod
Posts: 6718
Joined: Wed May 18, 2022 12:42 pm

Re: Pre-tax vs Roth for HNW

Post by toddthebod »

Luckywon wrote: Sun Jun 09, 2024 6:39 pm
toddthebod wrote: Sun Jun 09, 2024 2:08 pm
Luckywon wrote: Sun Jun 09, 2024 2:05 pm
toddthebod wrote: Sun Jun 09, 2024 1:15 pm That is not how the IRD deduction is calculated. Based on my rudimentary understanding, I think it works as follows: you calculate the estate tax twice, once including the IRA, once without, and the difference is the deductible amount. In this case, since $15M - $3M is less than the annual exclusion, the entire $832,000 should be deductible.
Ahh that sheds a lot of light. I knew I must be missing something as bsteiner is always correct!! :beer
It looks way more complicated in reality, but I'm pretty sure the basic idea is right, that the deduction for the estate tax allocated to the IRA is based on the estate tax calculation with and without that IRA.
It seems then a way of looking at this is that the tax deferred portion of one's estate is effectively not subject to estate tax?
If I am understanding this correctly, what happens is that the estate tax paid on the tax-deferred portion of the income is deducted from the income of the beneficiary, it's not a tax credit. I'll have to learn more about this and consider the consequences.
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

toddthebod wrote: Sun Jun 09, 2024 6:56 pm
Luckywon wrote: Sun Jun 09, 2024 6:39 pm
toddthebod wrote: Sun Jun 09, 2024 2:08 pm
Luckywon wrote: Sun Jun 09, 2024 2:05 pm
toddthebod wrote: Sun Jun 09, 2024 1:15 pm That is not how the IRD deduction is calculated. Based on my rudimentary understanding, I think it works as follows: you calculate the estate tax twice, once including the IRA, once without, and the difference is the deductible amount. In this case, since $15M - $3M is less than the annual exclusion, the entire $832,000 should be deductible.
Ahh that sheds a lot of light. I knew I must be missing something as bsteiner is always correct!! :beer
It looks way more complicated in reality, but I'm pretty sure the basic idea is right, that the deduction for the estate tax allocated to the IRA is based on the estate tax calculation with and without that IRA.
It seems then a way of looking at this is that the tax deferred portion of one's estate is effectively not subject to estate tax?
If I am understanding this correctly, what happens is that the estate tax paid on the tax-deferred portion of the income is deducted from the income of the beneficiary, it's not a tax credit. I'll have to learn more about this and consider the consequences.
Perhaps this will help on the Federal level....
https://smartasset.com/taxes/all-about-the-estate-tax
toddthebod
Posts: 6718
Joined: Wed May 18, 2022 12:42 pm

Re: Pre-tax vs Roth for HNW

Post by toddthebod »

Okay, I had to work through an example to really wrap my head around this.

Imagine a parent with a large estate above the exclusion that we will hold fixed in our two cases and ignore. In addition, we have $2,000,000 in cash and $1,000,000 in a traditional IRA.

Case 1: Let the beneficiary inherit the traditional IRA:

$1,000,000 in a traditional IRA plus $2,000,000 in cash results in $1,200,000 in estate taxes, paid out of cash.

Beneficiary inherits $1,000,000 traditional IRA and $800,000 in cash. He then liquidates the IRA, but due to the deduction for estate tax only pays tax on $600,000 of income. At the 37% bracket, that is $222,000, leaving him with an after-tax total of $1,578,000.

Case 2: Parent converts to Roth before he dies.

Parent pays 37% tax rate on conversion out of cash ($370,000) to leave the IRA as large as possible, leaving a $1,000,000 Roth IRA plus $1,630,000 in cash. This results in estate taxes of $1,052,000. The estate taxes are paid out of cash, leaving a $1,000,000 Roth IRA plus $578,000 in cash, the same after-tax total as in case 1.
stay.the.course
Posts: 17
Joined: Thu Apr 04, 2024 4:29 am

Re: Pre-tax vs Roth for HNW

Post by stay.the.course »

Estate tax exemption is around $27 mil for a married couple.

Let’s assume They both have $30 mil.

Husband passes. The estate moves to his/her spouse tax free.

The spouse passes. Now, estate tax kicks in.

Anything above $27 mil is subject to estate tax.

In anticipation of being $3 mil above the exemption, the wife converts $8 mil from tax deferred to Roth. She pays the taxes from a brocherage account (37% or $3 mil).

She has essentially decreased her estate to $27 mil by this massive conversion. Estate taxes due: $0. Kids will not pay inheritance tax (different than estate tax) on the $8 mil because it passes to them as Roth.

Scenario 2:

Wife decides not to convert. Estate taxes owed on $3 million ($1.2 million) and kid has to pay ordinary income tax on $8 million ( $3 million, 37% tax bracket). Total tax liability: $4.2 million.

Looks like by initiating the Roth conversion, she saved her heirs $1.2 million in taxes, and allowed the $8 million to grow to $16 mil 10 years later tax free!!!

Question: when someone inherits a Roth, does the growth of the account after it transfers to the heir continue to be tax free? In other words, if I empty it 10 years later and it has doubled, do I take the basis and the growth tax free?

If the answer is yes (an inherited Roth can grow tax free for 10 years and then cashed out tax free), then every person above the estate tax exception should do this (provided they have and want to pass things on to kids).
Luckywon
Posts: 2422
Joined: Tue Mar 28, 2017 10:33 am

Re: Pre-tax vs Roth for HNW

Post by Luckywon »

toddthebod wrote: Sun Jun 09, 2024 8:39 pm Okay, I had to work through an example to really wrap my head around this.

Imagine a parent with a large estate above the exclusion that we will hold fixed in our two cases and ignore. In addition, we have $2,000,000 in cash and $1,000,000 in a traditional IRA.

Case 1: Let the beneficiary inherit the traditional IRA:

$1,000,000 in a traditional IRA plus $2,000,000 in cash results in $1,200,000 in estate taxes, paid out of cash.

Beneficiary inherits $1,000,000 traditional IRA and $800,000 in cash. He then liquidates the IRA, but due to the deduction for estate tax only pays tax on $600,000 of income. At the 37% bracket, that is $222,000, leaving him with an after-tax total of $1,578,000.

Case 2: Parent converts to Roth before he dies.

Parent pays 37% tax rate on conversion out of cash ($370,000) to leave the IRA as large as possible, leaving a $1,000,000 Roth IRA plus $1,630,000 in cash. This results in estate taxes of $1,052,000. The estate taxes are paid out of cash, leaving a $1,000,000 Roth IRA plus $578,000 in cash, the same after-tax total as in case 1.
Thanks for this, very clarifying.

So though post tax dollars are the same, the conversion strategy does result in the converted dollars in a Roth account rather than a regular after tax account. The conversion appears to be functionally like a Mega large backdoor Roth, with the advantage of tax free growth/dividends over the applicable distribution period of the Roth.

That's if the parent died the day following the conversion. In the almost certain alternative, the effect of prepaying the tax and having less growth in the cash account during the lifetime of the parent should be factored in and would muddy the waters further? Just continuing the thought exercise as best I can, though at this point the complexity is already beyond my limit.
Zedon
Posts: 77
Joined: Thu Nov 06, 2014 1:37 am

Re: Pre-tax vs Roth for HNW

Post by Zedon »

I would think you are going to get in the highest tax bracket pretty easily, a maxed out Roth is worth way more than a maxed out traditional.
User avatar
illumination
Posts: 3297
Joined: Tue Apr 02, 2019 6:13 pm

Re: Pre-tax vs Roth for HNW

Post by illumination »

For the sake of argument, let's say the IRD deduction cancels out the "taxed twice" scenario of dying with both an estate tax due and a pre-tax IRA also having taxes owed. So it's tax neutral on the extra amount of estate tax it produces (still doubtful about this, would love an authoritative source, but let's pretend it does). So the 40% instant estate tax hit basically gets nullified as the pre-tax account gets drawn down in the form of an IRD deduction to the beneficiaries. Even in this scenario, it seems you're paying the estate tax upfront and then getting it back through a deduction. Usually the opposite of the strategy of deferring taxes.

But even with that, it still seems to ignore the bonus of leaving the heirs a Roth IRA for the next 10 years that they can collect at the end tax free. I feel like that part of the equation is being left out and it's a significant piece. Particularly if the heirs are likely going to be in a higher tax bracket while the Traditional IRA is being emptied out.

For the record, my estate attorney told me it was a good idea to make large conversions to Roth when the estate tax comes into play. He could be wrong, but it's sounds logical from my understanding. Would love to know the reasons why it would be a bad strategy.
toddthebod
Posts: 6718
Joined: Wed May 18, 2022 12:42 pm

Re: Pre-tax vs Roth for HNW

Post by toddthebod »

illumination wrote: Mon Jun 10, 2024 12:33 pm For the sake of argument, let's say the IRD deduction cancels out the "taxed twice" scenario of dying with both an estate tax due and a pre-tax IRA also having taxes owed. So it's tax neutral (still doubtful about this, would love an authoritative source, but let's pretend it does). So the 40% instant estate tax hit basically gets nullified as pre-tax account gets drawn down in the form of an IRD deduction to the beneficiaries.

That still seems to ignore the bonus of leaving the heirs a Roth IRA for the next 10 years that they can collect at the end tax free. I feel like that part of the equation is being left out and it's a significant piece. Particularly if the heirs are likely going to be in a higher tax bracket while the Traditional IRA is being emptied out.

For the record, my estate attorney told me it was a good idea to make large conversions to Roth when the estate tax comes into play. He could be wrong, but it's sounds logical from my understanding. Would love to know the reasons why it would be a bad strategy.
If you and your heir are in the same tax bracket or your heir is in a higher bracket, Roth conversions before passing are a no brainier. Someone with a taxable estate, however, might be in a very high bracket just from RMDs, and the question is whether it is better to convert at 32-37% if your heirs are in the 22-24% bracket or even lower.
User avatar
illumination
Posts: 3297
Joined: Tue Apr 02, 2019 6:13 pm

Re: Pre-tax vs Roth for HNW

Post by illumination »

toddthebod wrote: Mon Jun 10, 2024 1:17 pm
illumination wrote: Mon Jun 10, 2024 12:33 pm For the sake of argument, let's say the IRD deduction cancels out the "taxed twice" scenario of dying with both an estate tax due and a pre-tax IRA also having taxes owed. So it's tax neutral (still doubtful about this, would love an authoritative source, but let's pretend it does). So the 40% instant estate tax hit basically gets nullified as pre-tax account gets drawn down in the form of an IRD deduction to the beneficiaries.

That still seems to ignore the bonus of leaving the heirs a Roth IRA for the next 10 years that they can collect at the end tax free. I feel like that part of the equation is being left out and it's a significant piece. Particularly if the heirs are likely going to be in a higher tax bracket while the Traditional IRA is being emptied out.

For the record, my estate attorney told me it was a good idea to make large conversions to Roth when the estate tax comes into play. He could be wrong, but it's sounds logical from my understanding. Would love to know the reasons why it would be a bad strategy.
If you and your heir are in the same tax bracket or your heir is in a higher bracket, Roth conversions before passing are a no brainier. Someone with a taxable estate, however, might be in a very high bracket just from RMDs, and the question is whether it is better to convert at 32-37% if your heirs are in the 22-24% bracket or even lower.


But you would also have to take into account, a large pre-tax account is also going to produce giant RMDs that likely make the beneficiaries tax bracket also jump up. Particularly over a compressed, 10-year time period. There will be some discount from the IRD deduction to the beneficiaries, but that doesnt cover the entire RMD amount.

I'm sure there are unique scenarios where it's maybe it's better take the estate tax hit and the IRD deduction, but it seems more of a rare case and has a lot of unique variables. I would tend to default to converting a pre-tax to a Roth if I felt for certain I was subject to an Estate Tax. I'm sort of surprised to see some people say the strategy has no value.
EricGold
Posts: 653
Joined: Sat Mar 16, 2024 4:19 pm

Re: Pre-tax vs Roth for HNW

Post by EricGold »

smitcat wrote: Sat Jun 08, 2024 10:59 am Both RPM and Pralana can handle these for you easily.
I favor Pralana as it has options to 'automatically' optimize Roth converrsions and drawdowns or make it manual.
I'm waiting for the online version of Pralana, specifically to help with this question of Roth conversions in the context of future RMD and IRMAA. My home-brew spreadsheet attempts do not inspire confidence.

As an aside, I've warmed up to the idea that as a simplifying method, I can ignore IRMAA. By that I mean that bunching income (via Roth conversions) does not seem to change average long-term IRMAA by much
User avatar
TomatoTomahto
Posts: 17549
Joined: Mon Apr 11, 2011 1:48 pm

Re: Pre-tax vs Roth for HNW

Post by TomatoTomahto »

toddthebod wrote: Mon Jun 10, 2024 1:17 pm
If you and your heir are in the same tax bracket or your heir is in a higher bracket, Roth conversions before passing are a no brainier. Someone with a taxable estate, however, might be in a very high bracket just from RMDs, and the question is whether it is better to convert at 32-37% if your heirs are in the 22-24% bracket or even lower.
1 of our 4 heirs is in the highest tax bracket as are we. In addition to 3 of our heirs, charoties would be better served by not converting.

Not Roth converting leaves options open: you can be more charitable than you had expected, tax laws might change and you can decide how to approach it, you or heirs might move out of the USA and weigh tax treaties, you can convert later. But, once converted you can’t go back and convert to pre-tax.

It’s possible to be in a high tax bracket due to RMDs, but I think it’s rare (other than some edge cases) to have more than $5M in pre-tax per person (absent inheritanced IRAs).
I get the FI part but not the RE part of FIRE.
User avatar
FiveK
Posts: 16174
Joined: Sun Mar 16, 2014 2:43 pm

Re: Pre-tax vs Roth for HNW

Post by FiveK »

EricGold wrote: Mon Jun 10, 2024 2:33 pm
smitcat wrote: Sat Jun 08, 2024 10:59 am Both RPM and Pralana can handle these for you easily.
I favor Pralana as it has options to 'automatically' optimize Roth converrsions and drawdowns or make it manual.
I'm waiting for the online version of Pralana, specifically to help with this question of Roth conversions in the context of future RMD and IRMAA. My home-brew spreadsheet attempts do not inspire confidence.
Meanwhile, a good approach is
a) estimate your marginal tax when taking both SS and RMDs
b) convert through that marginal tax rate now (assuming that conversion amount doesn't change the estimate in step "a").

You might reach a different answer using a more complicated approach, but even if the more complicated answer is correct it is unlikely to make a significant difference in your retirement quality of life. And there remains the chance that the more complicated answer may not be correct (e.g., the problem of assuming higher returns in a Roth account leading to huge suggested Roth conversions).
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

EricGold wrote: Mon Jun 10, 2024 2:33 pm
smitcat wrote: Sat Jun 08, 2024 10:59 am Both RPM and Pralana can handle these for you easily.
I favor Pralana as it has options to 'automatically' optimize Roth converrsions and drawdowns or make it manual.
I'm waiting for the online version of Pralana, specifically to help with this question of Roth conversions in the context of future RMD and IRMAA. My home-brew spreadsheet attempts do not inspire confidence.

As an aside, I've warmed up to the idea that as a simplifying method, I can ignore IRMAA. By that I mean that bunching income (via Roth conversions) does not seem to change average long-term IRMAA by much
We are not waiting for the new version of Pralana as this one does everything we need these past 5 years or so.
When the new web versions is available ....
- we will get it at a discount
- the data will be easily transferable
- this is the time of year we want to compare fwd models (not later)
- we will have our results completed to compare with the new version of Pralana

FWIW - in our case we are not ignoring IRMMA or any other costs. But that is just one of the things that Pralana has helped us with these past years.
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

TomatoTomahto wrote: Mon Jun 10, 2024 2:46 pm
toddthebod wrote: Mon Jun 10, 2024 1:17 pm
If you and your heir are in the same tax bracket or your heir is in a higher bracket, Roth conversions before passing are a no brainier. Someone with a taxable estate, however, might be in a very high bracket just from RMDs, and the question is whether it is better to convert at 32-37% if your heirs are in the 22-24% bracket or even lower.
1 of our 4 heirs is in the highest tax bracket as are we. In addition to 3 of our heirs, charoties would be better served by not converting.

Not Roth converting leaves options open: you can be more charitable than you had expected, tax laws might change and you can decide how to approach it, you or heirs might move out of the USA and weigh tax treaties, you can convert later. But, once converted you can’t go back and convert to pre-tax.

It’s possible to be in a high tax bracket due to RMDs, but I think it’s rare (other than some edge cases) to have more than $5M in pre-tax per person (absent inheritanced IRAs).
Typically folks convert when they are retired but have not yet elected SS payments - that leave their taxes more flexible ane lower to make the conversions more attractive.
In those cases conversions prior to electing SS is an option which closes to some degree and becomes mostly unavailable later on.
Once retired tax brackets are typically more controlable save for required RMD's which are proportional to the unconverted funds.
User avatar
TomatoTomahto
Posts: 17549
Joined: Mon Apr 11, 2011 1:48 pm

Re: Pre-tax vs Roth for HNW

Post by TomatoTomahto »

smitcat wrote: Mon Jun 10, 2024 4:13 pm
Typically folks convert when they are retired but have not yet elected SS payments - that leave their taxes more flexible ane lower to make the conversions more attractive.
In those cases conversions prior to electing SS is an option which closes to some degree and becomes mostly unavailable later on.
Once retired tax brackets are typically more controlable save for required RMD's which are proportional to the unconverted funds.
I sort of agree, but “typically folks …” doesn’t necessarily apply to HNW individuals, the focus of this thread.
I get the FI part but not the RE part of FIRE.
User avatar
illumination
Posts: 3297
Joined: Tue Apr 02, 2019 6:13 pm

Re: Pre-tax vs Roth for HNW

Post by illumination »

TomatoTomahto wrote: Mon Jun 10, 2024 2:46 pm
toddthebod wrote: Mon Jun 10, 2024 1:17 pm
If you and your heir are in the same tax bracket or your heir is in a higher bracket, Roth conversions before passing are a no brainier. Someone with a taxable estate, however, might be in a very high bracket just from RMDs, and the question is whether it is better to convert at 32-37% if your heirs are in the 22-24% bracket or even lower.
1 of our 4 heirs is in the highest tax bracket as are we. In addition to 3 of our heirs, charoties would be better served by not converting.

Not Roth converting leaves options open: you can be more charitable than you had expected, tax laws might change and you can decide how to approach it, you or heirs might move out of the USA and weigh tax treaties, you can convert later. But, once converted you can’t go back and convert to pre-tax.



If I was hoping for my kids to give the money to charity...I would just go ahead and give the money to charity.

I wouldnt want to strategically put them in a position where they face high taxes and feel pressured to give more than they would on their own accord in order to chase a tax break.

I feel "not converting" closes more door than it opens. You can't go back and convert if you made the wrong choice, but you always have the ability to be generous and give away more of your money.

TomatoTomahto wrote: Mon Jun 10, 2024 2:46 pm
It’s possible to be in a high tax bracket due to RMDs, but I think it’s rare (other than some edge cases) to have more than $5M in pre-tax per person (absent inheritanced IRAs).
It's actually pretty easy (imo) for a large IRA to push people into the higher brackets. Especially since when the "parents" die, very likely for the adult kids to be near the peak earning years of their career. And the RMDs for an inherited IRA are on a 10 year, compressed schedule, and many will be taking a large distribution every year. Maybe a married couple is making $200k in their 50's (in many parts of the country this is relatively middle class now) they inherit a $3 million IRA and they take out $300k+ a year, you're now in the 2nd highest tax bracket for the next decade with $500k a year. Throw some reasonable assumptions about compounding and it gets "worse". There's various deductions, but you can see it's not that crazy of a scenario to where it can be a big jump. You also have to take in consideration we're discussing cases where estates are surpassing the estate tax threshold, so already large (edge) cases. Very unlikely an heir goes from $0 in earnings to a large RMD overnight.

-
EricGold
Posts: 653
Joined: Sat Mar 16, 2024 4:19 pm

Re: Pre-tax vs Roth for HNW

Post by EricGold »

FiveK wrote: Mon Jun 10, 2024 3:14 pm
EricGold wrote: Mon Jun 10, 2024 2:33 pm I'm waiting for the online version of Pralana, specifically to help with this question of Roth conversions in the context of future RMD and IRMAA. My home-brew spreadsheet attempts do not inspire confidence.
Meanwhile, a good approach is
a) estimate your marginal tax when taking both SS and RMDs
b) convert through that marginal tax rate now (assuming that conversion amount doesn't change the estimate in step "a").
Thanks for the suggestion
How would you handle the case of increasing tax brackets over time with, or without, Roth conversion ? This is where my calculations are not up to the task.
User avatar
TomatoTomahto
Posts: 17549
Joined: Mon Apr 11, 2011 1:48 pm

Re: Pre-tax vs Roth for HNW

Post by TomatoTomahto »

illumination wrote: Mon Jun 10, 2024 5:00 pm If I was hoping for my kids to give the money to charity...I would just go ahead and give the money to charity.

I wouldnt want to strategically put them in a position where they face high taxes and feel pressured to give more than they would on their own accord in order to chase a tax break.
I guess I didn’t express myself well. I wasn’t speaking of my kids. I had no idea that I would want to donate as much to charity as I now do. Doing so with pre-tax money is wonderfully efficient. I wish that I had not contributed/converted to Roth, but it is what it is.
I get the FI part but not the RE part of FIRE.
fyre4ce
Posts: 2616
Joined: Sun Aug 06, 2017 11:29 am

Re: Pre-tax vs Roth for HNW

Post by fyre4ce »

illumination wrote: Mon Jun 10, 2024 12:33 pm For the sake of argument, let's say the IRD deduction cancels out the "taxed twice" scenario of dying with both an estate tax due and a pre-tax IRA also having taxes owed. So it's tax neutral on the extra amount of estate tax it produces (still doubtful about this, would love an authoritative source, but let's pretend it does). So the 40% instant estate tax hit basically gets nullified as the pre-tax account gets drawn down in the form of an IRD deduction to the beneficiaries. Even in this scenario, it seems you're paying the estate tax upfront and then getting it back through a deduction. Usually the opposite of the strategy of deferring taxes.

But even with that, it still seems to ignore the bonus of leaving the heirs a Roth IRA for the next 10 years that they can collect at the end tax free. I feel like that part of the equation is being left out and it's a significant piece. Particularly if the heirs are likely going to be in a higher tax bracket while the Traditional IRA is being emptied out.

For the record, my estate attorney told me it was a good idea to make large conversions to Roth when the estate tax comes into play. He could be wrong, but it's sounds logical from my understanding. Would love to know the reasons why it would be a bad strategy.
My understanding is that the IRD deduction equalizes estate and income taxes for Roth conversions on estates subject to the estate tax. I believe it works like this:

Say you have a $1M pre-tax account and a $2M taxable account, all estate-taxable at 40% due to other assets in the estate. Both you and your heir are in the 37% federal income tax bracket.

With no IRD deduction, Roth conversion would be an advantage. With no conversion, your estate would pay $1.2M (=$3Mx40%) estate taxes from the taxable account, then your heir would pay $370k (=$1Mx37%) income taxes on the pre-tax assets, and be left with $1.43M. But if you converted the pre-tax money to Roth, you would pay $370k income tax, then your estate would only pay $1.052M (=$2.63Mx40%), leaving your heir with $1.578M in all after-tax assets, a $148k improvement.

The IRD deduction gets taken on the heir's personal income tax return, and the amount is equal to the estate taxes paid on the pre-tax assets they inherit, which in this case is $400k (=$1Mx40%). So if you did no Roth conversion, your estate would still pay $1.2M in estate tax, but then your heir would only pay income tax on $600k of income (=$1M-$400k), giving an income tax of $222k (=$600k x 37%). This would leave them with $1.578M after-tax (=$3M-$1.2M-$222k), exactly the same amount as if you did the Roth conversion. The IRD deduction equalizes estate and income taxes when income tax rates are equal.

One caution is the heir has to know to take it. There's almost zero chance someone's random heir will, unless specifically told by a tax pro. Likely their accountant isn't the same one who prepared the estate tax return, so will they know to ask if estate taxes were paid given how rare they are? Likely not, I would guess. So for anyone in this situation, I'd recommend making sure YOUR tax pro will reach out to heirs after your death and make sure they know to take the deduction.

The second caution is that the deduction is below-the-line on Schedule A "Other Itemized Deductions". There is no floor or cap, but this deduction won't lower AGI, which could affect phase-outs of credits, IRMAA, etc. And, any deduction needed for Schedule A to exceed the standard deduction gets "lost". So it might not fully equalize taxes.

There is no IRD deduction for state estate taxes. If facing state estate taxes, Roth conversions will be beneficial when [marginal tax rate of heir] > [your marginal income tax rate] x (1 - [marginal estate tax rate on your estate]). (This formula is derived on the wiki here.) If you have a 37% income tax rate, your heir has a 32% income tax rate, and you expect to pay a 16% marginal state estate tax rate, Roth conversions would still be beneficial (37% x (1 - 16%) = 31.08% which is less than 32%).

Roth conversions might still be beneficial for other reasons:

First, if your marginal income tax rate is less than your heirs (you might have lower income, live in a lower tax state, etc.) then you are getting a tax rate arbitrage.

Second, there is the "contributing the maximum" effect where Roth conversions with taxes paid from other assets get tax-protected growth on more money. That will be noticeable over the ten year stretch plus any years remaining in your own life. The formula is derived here. Over 10 years with a stock-like investment (8% return and 2% mostly-qualified yield), Roth conversions at 37% for heirs in the 32% bracket will still be beneficial (break-even at 31.8%).

Third, if leaving large pre-tax accounts per-heir, large withdrawals might artificially raise their tax rates. If you and your heir are both in the 32% bracket but you leave them a $2M pre-tax IRA, big chunks of their ~$200k/year withdrawals will likely be taxed at more than 32%. If you can't spread the pre-tax assets around to other heirs, Roth conversions during your lifetime will probably be beneficial. This won't happen if the heir is already in the 37% bracket. There's no way their rate can go higher, so they are best off leaving the money in until the end of the tenth year for maximum tax-protected growth, and that equalizes this effect when compared to Roth.
User avatar
FiveK
Posts: 16174
Joined: Sun Mar 16, 2014 2:43 pm

Re: Pre-tax vs Roth for HNW

Post by FiveK »

EricGold wrote: Mon Jun 10, 2024 5:11 pm
FiveK wrote: Mon Jun 10, 2024 3:14 pm
EricGold wrote: Mon Jun 10, 2024 2:33 pm I'm waiting for the online version of Pralana, specifically to help with this question of Roth conversions in the context of future RMD and IRMAA. My home-brew spreadsheet attempts do not inspire confidence.
Meanwhile, a good approach is
a) estimate your marginal tax when taking both SS and RMDs
b) convert through that marginal tax rate now (assuming that conversion amount doesn't change the estimate in step "a").
Thanks for the suggestion
How would you handle the case of increasing tax brackets over time with, or without, Roth conversion ? This is where my calculations are not up to the task.
If you don't convert, RMDs will get taxed at those increasing rates, leaving less after-tax spendable.

If you do convert, you can convert enough to keep the tax rate ~constant, and that should give you ~the most after-tax spendable income.
User avatar
illumination
Posts: 3297
Joined: Tue Apr 02, 2019 6:13 pm

Re: Pre-tax vs Roth for HNW

Post by illumination »

fyre4ce wrote: Mon Jun 10, 2024 5:35 pm
The IRD deduction gets taken on the heir's personal income tax return, and the amount is equal to the estate taxes paid on the pre-tax assets they inherit, which in this case is $400k (=$1Mx40%). So if you did no Roth conversion, your estate would still pay $1.2M in estate tax, but then your heir would only pay income tax on $600k of income (=$1M-$400k), giving an income tax of $222k (=$600k x 37%). This would leave them with $1.578M after-tax (=$3M-$1.2M-$222k), exactly the same amount as if you did the Roth conversion. The IRD deduction equalizes estate and income taxes when income tax rates are equal.

And I understand the point being made here, basically the extra estate tax generated from a pre-tax retirement plan being included in the estate can then be deducted by the heir against those pre-tax distributions. So it balances out. Which seems fair as you start getting into cases where you come close to 100% taxation on retirement accounts. Not that long ago the estate tax was 55% and the top marginal rate was 39.6%. Plus state taxes.

But that isn't the end of the upsides to convert to Roth for estate planning. If say the Roth IRA had to be immediately liquidated upon death (even with no tax being owed) than I could understand the IRD deduction wiping out most of the case to convert to a Roth.

What seems to keep getting missed though is the 10 years of tax free growth the heir will get after death.


Also, there's a time value of money aspect if you decide to not convert a pre-tax account as you're "pre-paying" a large estate tax and then filing to get the tax you already paid deducted over the next 10 years. The opposite of tax deferral. And does this deduction go away with tax code changes? Probably the group the public has the least "sympathy" for are those who inherited large sums of money and want to deduct the estate tax on their large RMDs.
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

FiveK wrote: Mon Jun 10, 2024 5:52 pm
EricGold wrote: Mon Jun 10, 2024 5:11 pm
FiveK wrote: Mon Jun 10, 2024 3:14 pm
EricGold wrote: Mon Jun 10, 2024 2:33 pm I'm waiting for the online version of Pralana, specifically to help with this question of Roth conversions in the context of future RMD and IRMAA. My home-brew spreadsheet attempts do not inspire confidence.
Meanwhile, a good approach is
a) estimate your marginal tax when taking both SS and RMDs
b) convert through that marginal tax rate now (assuming that conversion amount doesn't change the estimate in step "a").
Thanks for the suggestion
How would you handle the case of increasing tax brackets over time with, or without, Roth conversion ? This is where my calculations are not up to the task.
If you don't convert, RMDs will get taxed at those increasing rates, leaving less after-tax spendable.

If you do convert, you can convert enough to keep the tax rate ~constant, and that should give you ~the most after-tax spendable income.
"If you do convert, you can convert enough to keep the tax rate ~constant, and that should give you ~the most after-tax spendable income."
I think this requires calculating the tax deferred accounts entire lifetime with your best estimates of both your porfolio performance and your draw rates for each year without inflation. Using a baseline of no conversions and then a camparison with conversions. It would be best to also ensure that your drawdown strategy was optimized and perhaps test alternate future scenarios including but not limted to: age of demise of 1st spouse, alternate levels of spending, alternate portfolio performance %'s.
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

TomatoTomahto wrote: Mon Jun 10, 2024 4:39 pm
smitcat wrote: Mon Jun 10, 2024 4:13 pm
Typically folks convert when they are retired but have not yet elected SS payments - that leave their taxes more flexible ane lower to make the conversions more attractive.
In those cases conversions prior to electing SS is an option which closes to some degree and becomes mostly unavailable later on.
Once retired tax brackets are typically more controlable save for required RMD's which are proportional to the unconverted funds.
I sort of agree, but “typically folks …” doesn’t necessarily apply to HNW individuals, the focus of this thread.
I thought the OP was looking at comparing pre-tax and Roth while still working for HNW folks. Many of them that I know certianly did not continue work at later ages after achieving their higher net worth. That is the basic strategy for Roth conversions.
User avatar
FiveK
Posts: 16174
Joined: Sun Mar 16, 2014 2:43 pm

Re: Pre-tax vs Roth for HNW

Post by FiveK »

smitcat wrote: Mon Jun 10, 2024 6:30 pm
FiveK wrote: Mon Jun 10, 2024 5:52 pm If you don't convert, RMDs will get taxed at those increasing rates, leaving less after-tax spendable.

If you do convert, you can convert enough to keep the tax rate ~constant, and that should give you ~the most after-tax spendable income.
I think this requires calculating the tax deferred accounts entire lifetime with your best estimates of both your porfolio performance and your draw rates for each year without inflation. Using a baseline of no conversions and then a camparison with conversions. It would be best to also ensure that your drawdown strategy was optimized and perhaps test alternate future scenarios including but not limted to: age of demise of 1st spouse, alternate levels of spending, alternate portfolio performance %'s.
One can certainly test a plethora of scenarios, but in the end one has to choose a specific action for "what do I do this year?" E.g., there are no do-overs on Roth conversions if one assumed the wrong end of the $17K/yr vs. $151K/yr spread exhibited in the Simple method section. And that doesn't even consider possible tax law changes. You take your best guess, and what happens happens.

Numerical optimization is also difficult when the functions under study look like the chart in Worth pushing through the Social Security hump and/or IRMAA cliffs?.

Creating complex models and dealing with complex optimization can be fun - for those so inclined. ;) Any learning from analysis of such models and their results can feed back into simpler strategies. Thus there are many reasons to support doing so for those who enjoy it.

For the larger fraction of the population, a simple "this year plan" based on a rough comparison of current vs. future conditions is likely to serve them well.
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

FiveK wrote: Mon Jun 10, 2024 7:03 pm
smitcat wrote: Mon Jun 10, 2024 6:30 pm
FiveK wrote: Mon Jun 10, 2024 5:52 pm If you don't convert, RMDs will get taxed at those increasing rates, leaving less after-tax spendable.

If you do convert, you can convert enough to keep the tax rate ~constant, and that should give you ~the most after-tax spendable income.
I think this requires calculating the tax deferred accounts entire lifetime with your best estimates of both your porfolio performance and your draw rates for each year without inflation. Using a baseline of no conversions and then a camparison with conversions. It would be best to also ensure that your drawdown strategy was optimized and perhaps test alternate future scenarios including but not limted to: age of demise of 1st spouse, alternate levels of spending, alternate portfolio performance %'s.
One can certainly test a plethora of scenarios, but in the end one has to choose a specific action for "what do I do this year?" E.g., there are no do-overs on Roth conversions if one assumed the wrong end of the $17K/yr vs. $151K/yr spread exhibited in the Simple method section. And that doesn't even consider possible tax law changes. You take your best guess, and what happens happens.

Numerical optimization is also difficult when the functions under study look like the chart in Worth pushing through the Social Security hump and/or IRMAA cliffs?.

Creating complex models and dealing with complex optimization can be fun - for those so inclined. ;) Any learning from analysis of such models and their results can feed back into simpler strategies. Thus there are many reasons to support doing so for those who enjoy it.

For the larger fraction of the population, a simple "this year plan" based on a rough comparison of current vs. future conditions is likely to serve them well.
"For the larger fraction of the population, a simple "this year plan" based on a rough comparison of current vs. future conditions is likely to serve them well."
How do yo do this in the simplest form without using mutiple years for a comparison?
How do you configure the 'this year plan" vs 'future plan'?
User avatar
TomatoTomahto
Posts: 17549
Joined: Mon Apr 11, 2011 1:48 pm

Re: Pre-tax vs Roth for HNW

Post by TomatoTomahto »

smitcat wrote: Mon Jun 10, 2024 6:35 pm
TomatoTomahto wrote: Mon Jun 10, 2024 4:39 pm
smitcat wrote: Mon Jun 10, 2024 4:13 pm
Typically folks convert when they are retired but have not yet elected SS payments - that leave their taxes more flexible ane lower to make the conversions more attractive.
In those cases conversions prior to electing SS is an option which closes to some degree and becomes mostly unavailable later on.
Once retired tax brackets are typically more controlable save for required RMD's which are proportional to the unconverted funds.
I sort of agree, but “typically folks …” doesn’t necessarily apply to HNW individuals, the focus of this thread.
I thought the OP was looking at comparing pre-tax and Roth while still working for HNW folks. Many of them that I know certianly did not continue work at later ages after achieving their higher net worth. That is the basic strategy for Roth conversions.
I think I lost the plot somewhere in this thread. For better or worse, we are HNW, our marginal tax rate is high (46%), my wife is still working, and while she was doing Roth 401k for a while, she stopped doing so in favor of traditional. She chooses to continue to work and there will probably be no time where our marginal rate is lower than 42% before RMDs begin for her (they have already begun for me).
I get the FI part but not the RE part of FIRE.
User avatar
FiveK
Posts: 16174
Joined: Sun Mar 16, 2014 2:43 pm

Re: Pre-tax vs Roth for HNW

Post by FiveK »

smitcat wrote: Mon Jun 10, 2024 7:13 pm
FiveK wrote: Mon Jun 10, 2024 7:03 pm
smitcat wrote: Mon Jun 10, 2024 6:30 pm I think this requires calculating the tax deferred accounts entire lifetime with your best estimates of both your porfolio performance and your draw rates for each year without inflation. Using a baseline of no conversions and then a camparison with conversions. It would be best to also ensure that your drawdown strategy was optimized and perhaps test alternate future scenarios including but not limted to: age of demise of 1st spouse, alternate levels of spending, alternate portfolio performance %'s.
One can certainly test a plethora of scenarios, but in the end one has to choose a specific action for "what do I do this year?" E.g., there are no do-overs on Roth conversions if one assumed the wrong end of the $17K/yr vs. $151K/yr spread exhibited in the Simple method section. And that doesn't even consider possible tax law changes. You take your best guess, and what happens happens.

Numerical optimization is also difficult when the functions under study look like the chart in Worth pushing through the Social Security hump and/or IRMAA cliffs?.

Creating complex models and dealing with complex optimization can be fun - for those so inclined. ;) Any learning from analysis of such models and their results can feed back into simpler strategies. Thus there are many reasons to support doing so for those who enjoy it.

For the larger fraction of the population, a simple "this year plan" based on a rough comparison of current vs. future conditions is likely to serve them well.
How do yo do this in the simplest form without using mutiple years for a comparison?
How do you configure the 'this year plan" vs 'future plan'?
1) Pick a representative year, such as "several years after starting RMDs" and calculate the marginal tax rate for that year using real returns, assumed withdrawal rate, and current tax law.
2) Determine the income for this year that will have this year's marginal tax rate match the one from step #1.
3) Repeat steps 1 & 2 next year.

Perfectly optimal? Probably not.
Close enough? Probably.

Perhaps the biggest "oh-by-the-way" is the extent to which "this year's" choice will affect the future. Given the ~$300K/yr range for MFJ filers that covers the 22% and 24% federal brackets, there is a lot of room for error for those with income in that range. :)
User avatar
vnatale
Posts: 3893
Joined: Sat Jul 31, 2010 8:50 pm
Location: Montague, MA

Re: Pre-tax vs Roth for HNW

Post by vnatale »

I'm a no debt person. Never had a car loan. Did not borrow to buy my cheap house.

Future tax liabilities on retirement investments are a form of debt.

Therefore, as soon as Roth because available in 1998 I converted ALL my investments.

I just wanted to be forever done with any future taxes on those investments.

Subsequently, if I could, I almost always chose the Roth option.

At the time in 1998 I was not aware of three other benefits:

1) Reduced Require Minimum Distributions

2) Better able to say out of the surcharges for Medicare premiums

3) Better able to avoid having the maximum - 85% - of Social Security income being subject to taxation.
Above provided by: Vinny, who always says: "I only regret that I have but one lap to give to my cats." AND "I'm a more-is-more person."
EricGold
Posts: 653
Joined: Sat Mar 16, 2024 4:19 pm

Re: Pre-tax vs Roth for HNW

Post by EricGold »

smitcat wrote: Mon Jun 10, 2024 6:30 pm
I think this requires calculating the tax deferred accounts entire lifetime with your best estimates of both your porfolio performance and your draw rates for each year without inflation. Using a baseline of no conversions and then a camparison with conversions.
Yep -- that is what I also concluded and tried to do. But it is tedious at best since I don't have an easy way to program a loop in a spreadsheet, and I'm feeling too lazy to learn enough JS again. I also want to model a reversion to circa 2017 tax brackets at different times in the future. Thus my waiting for Pralana online
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

TomatoTomahto wrote: Mon Jun 10, 2024 7:19 pm
smitcat wrote: Mon Jun 10, 2024 6:35 pm
TomatoTomahto wrote: Mon Jun 10, 2024 4:39 pm
smitcat wrote: Mon Jun 10, 2024 4:13 pm
Typically folks convert when they are retired but have not yet elected SS payments - that leave their taxes more flexible ane lower to make the conversions more attractive.
In those cases conversions prior to electing SS is an option which closes to some degree and becomes mostly unavailable later on.
Once retired tax brackets are typically more controlable save for required RMD's which are proportional to the unconverted funds.
I sort of agree, but “typically folks …” doesn’t necessarily apply to HNW individuals, the focus of this thread.
I thought the OP was looking at comparing pre-tax and Roth while still working for HNW folks. Many of them that I know certianly did not continue work at later ages after achieving their higher net worth. That is the basic strategy for Roth conversions.
I think I lost the plot somewhere in this thread. For better or worse, we are HNW, our marginal tax rate is high (46%), my wife is still working, and while she was doing Roth 401k for a while, she stopped doing so in favor of traditional. She chooses to continue to work and there will probably be no time where our marginal rate is lower than 42% before RMDs begin for her (they have already begun for me).
I thought the plot was discussing the typical way folks approach this issue not anything personal at all.
In most any article or paper the disussions about Roth accounts and Roth conversions use a strategy that targets the time period after someone has earned income from a job and when they start receiving fixed incomes from pensions/SS/etc. This is always discussed as the target time to adjust accounts utilizing Roth conversions as the best opportunity.
When folks are in this zone they are typically in a low tax bracket - even folks with 10"s of milions of $$$ in funds can easily be in the 24% bracket at that time which offers these opportunities.
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

FiveK wrote: Mon Jun 10, 2024 7:33 pm
smitcat wrote: Mon Jun 10, 2024 7:13 pm
FiveK wrote: Mon Jun 10, 2024 7:03 pm
smitcat wrote: Mon Jun 10, 2024 6:30 pm I think this requires calculating the tax deferred accounts entire lifetime with your best estimates of both your porfolio performance and your draw rates for each year without inflation. Using a baseline of no conversions and then a camparison with conversions. It would be best to also ensure that your drawdown strategy was optimized and perhaps test alternate future scenarios including but not limted to: age of demise of 1st spouse, alternate levels of spending, alternate portfolio performance %'s.
One can certainly test a plethora of scenarios, but in the end one has to choose a specific action for "what do I do this year?" E.g., there are no do-overs on Roth conversions if one assumed the wrong end of the $17K/yr vs. $151K/yr spread exhibited in the Simple method section. And that doesn't even consider possible tax law changes. You take your best guess, and what happens happens.

Numerical optimization is also difficult when the functions under study look like the chart in Worth pushing through the Social Security hump and/or IRMAA cliffs?.

Creating complex models and dealing with complex optimization can be fun - for those so inclined. ;) Any learning from analysis of such models and their results can feed back into simpler strategies. Thus there are many reasons to support doing so for those who enjoy it.

For the larger fraction of the population, a simple "this year plan" based on a rough comparison of current vs. future conditions is likely to serve them well.
How do yo do this in the simplest form without using mutiple years for a comparison?
How do you configure the 'this year plan" vs 'future plan'?
1) Pick a representative year, such as "several years after starting RMDs" and calculate the marginal tax rate for that year using real returns, assumed withdrawal rate, and current tax law.
2) Determine the income for this year that will have this year's marginal tax rate match the one from step #1.
3) Repeat steps 1 & 2 next year.

Perfectly optimal? Probably not.
Close enough? Probably.

Perhaps the biggest "oh-by-the-way" is the extent to which "this year's" choice will affect the future. Given the ~$300K/yr range for MFJ filers that covers the 22% and 24% federal brackets, there is a lot of room for error for those with income in that range. :)
"1) Pick a representative year, such as "several years after starting RMDs" and calculate the marginal tax rate for that year using real returns, assumed withdrawal rate, and current tax law.
2) Determine the income for this year that will have this year's marginal tax rate match the one from step #1.
3) Repeat steps 1 & 2 next year."

Correct me if I am wrong but this will only work if....
- the drawdown level is mostly keeping time with the rising RMD % (otherwise said - somwhere around a 4% draw)
- the persons portfolio AA's goals are to remain the same out to their demise (otherwose said - no large changes in AA over time)
- the person is not actively planning for heirs (charitable will show up in early years but heirs mostly show up at the end)
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

EricGold wrote: Mon Jun 10, 2024 8:46 pm
smitcat wrote: Mon Jun 10, 2024 6:30 pm
I think this requires calculating the tax deferred accounts entire lifetime with your best estimates of both your porfolio performance and your draw rates for each year without inflation. Using a baseline of no conversions and then a camparison with conversions.
Yep -- that is what I also concluded and tried to do. But it is tedious at best since I don't have an easy way to program a loop in a spreadsheet, and I'm feeling too lazy to learn enough JS again. I also want to model a reversion to circa 2017 tax brackets at different times in the future. Thus my waiting for Pralana online
It is easily done by Pralana and in most cases by RPM as well.
The outputs will include columns of numbers by year as well as totals for many things including: tax by year, total taxes,marginal rate, effective rate,tax deferred conversion by year, totals in each account, 'spendable' $$ by year, "Spendable" dollars total (plan), and so many others it is hard to list. These reports can be viewed and printed with or without inflation and can easily be compared, recorded and charted if desired for trends.
User avatar
TomatoTomahto
Posts: 17549
Joined: Mon Apr 11, 2011 1:48 pm

Re: Pre-tax vs Roth for HNW

Post by TomatoTomahto »

smitcat wrote: Tue Jun 11, 2024 7:32 am When folks are in this zone they are typically in a low tax bracket - even folks with 10"s of milions of $$$ in funds can easily be in the 24% bracket at that time which offers these opportunities.
Unless you have all 10s of millions of dollars in BRK, I think the 24% bracket is not “easily” reached. 2% is typical taxable income, then add in SS, any pensions, deferred compensation, etc.

We can agree to disagree and I’ll go bother someone else.
I get the FI part but not the RE part of FIRE.
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

TomatoTomahto wrote: Tue Jun 11, 2024 8:40 am
smitcat wrote: Tue Jun 11, 2024 7:32 am When folks are in this zone they are typically in a low tax bracket - even folks with 10"s of milions of $$$ in funds can easily be in the 24% bracket at that time which offers these opportunities.
Unless you have all 10s of millions of dollars in BRK, I think the 24% bracket is not “easily” reached. 2% is typical taxable income, then add in SS, any pensions, deferred compensation, etc.

We can agree to disagree and I’ll go bother someone else.

These are some paramaters to look at.
- $500K to no limit in Roth accounts
- a few million in municipal bonds
- $15 million or more in equity at say 1.5% Div average
- 401K up to $100K can be completely QCD's
- SS say $100K

$85K - SS taxable $85
$225K - 1.5% of 15 million
$310 total about $110 short of the 24% bracket.
A person could 'spend' $500K every year and never exceed the 24% bracket, a person could 'spend' a couple of million $$ in one year and never exceed the 24% bracket.
aristotelian
Posts: 12549
Joined: Wed Jan 11, 2017 7:05 pm

Re: Pre-tax vs Roth for HNW

Post by aristotelian »

smitcat wrote: Tue Jun 11, 2024 8:55 am
TomatoTomahto wrote: Tue Jun 11, 2024 8:40 am
smitcat wrote: Tue Jun 11, 2024 7:32 am When folks are in this zone they are typically in a low tax bracket - even folks with 10"s of milions of $$$ in funds can easily be in the 24% bracket at that time which offers these opportunities.
Unless you have all 10s of millions of dollars in BRK, I think the 24% bracket is not “easily” reached. 2% is typical taxable income, then add in SS, any pensions, deferred compensation, etc.

We can agree to disagree and I’ll go bother someone else.

These are some paramaters to look at.
- $500K to no limit in Roth accounts
- a few million in municipal bonds
- $15 million or more in equity at say 1.5% Div average
- 401K up to $100K can be completely QCD's
- SS say $100K

$85K - SS taxable $85
$225K - 1.5% of 15 million
$310 total about $110 short of the 24% bracket.
A person could 'spend' $500K every year and never exceed the 24% bracket, a person could 'spend' a couple of million $$ in one year and never exceed the 24% bracket.
The amount in taxable account is really not relevant to bracket for income tax as long as the positions are tax efficient and subject to rates for qualified dividends. With nothing but a standard Boglehead portfolio I would say it is actually pretty difficult to get into the 24% bracket. Where things might get difficult is if you have a pension, real estate income, W2 income, consulting income, etc. If your taxable Social Security offsets your standard deduction, you would still have room for close to $200k of Roth conversions, so you would need to have perhaps $2M in Traditional IRA before I would be worried about hitting the 24% bracket.

Where you could get into trouble is if you were planning to be in the 10-12% brackets which is the territory you could get a bigger ACA subsidy by keeping your AGI low. Then you get hit with the double whammy of losing the subsidy and increasing your tax bracket.
toddthebod
Posts: 6718
Joined: Wed May 18, 2022 12:42 pm

Re: Pre-tax vs Roth for HNW

Post by toddthebod »

smitcat wrote: Tue Jun 11, 2024 8:55 am
TomatoTomahto wrote: Tue Jun 11, 2024 8:40 am
smitcat wrote: Tue Jun 11, 2024 7:32 am When folks are in this zone they are typically in a low tax bracket - even folks with 10"s of milions of $$$ in funds can easily be in the 24% bracket at that time which offers these opportunities.
Unless you have all 10s of millions of dollars in BRK, I think the 24% bracket is not “easily” reached. 2% is typical taxable income, then add in SS, any pensions, deferred compensation, etc.

We can agree to disagree and I’ll go bother someone else.

These are some paramaters to look at.
- $500K to no limit in Roth accounts
- a few million in municipal bonds
- $15 million or more in equity at say 1.5% Div average
- 401K up to $100K can be completely QCD's
- SS say $100K

$85K - SS taxable $85
$225K - 1.5% of 15 million
$310 total about $110 short of the 24% bracket.
A person could 'spend' $500K every year and never exceed the 24% bracket, a person could 'spend' a couple of million $$ in one year and never exceed the 24% bracket.
Your dividends mostly have no effect on your tax bracket.
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

aristotelian wrote: Tue Jun 11, 2024 9:18 am
smitcat wrote: Tue Jun 11, 2024 8:55 am
TomatoTomahto wrote: Tue Jun 11, 2024 8:40 am
smitcat wrote: Tue Jun 11, 2024 7:32 am When folks are in this zone they are typically in a low tax bracket - even folks with 10"s of milions of $$$ in funds can easily be in the 24% bracket at that time which offers these opportunities.
Unless you have all 10s of millions of dollars in BRK, I think the 24% bracket is not “easily” reached. 2% is typical taxable income, then add in SS, any pensions, deferred compensation, etc.

We can agree to disagree and I’ll go bother someone else.

These are some paramaters to look at.
- $500K to no limit in Roth accounts
- a few million in municipal bonds
- $15 million or more in equity at say 1.5% Div average
- 401K up to $100K can be completely QCD's
- SS say $100K

$85K - SS taxable $85
$225K - 1.5% of 15 million
$310 total about $110 short of the 24% bracket.
A person could 'spend' $500K every year and never exceed the 24% bracket, a person could 'spend' a couple of million $$ in one year and never exceed the 24% bracket.
The amount in taxable account is really not relevant to bracket for income tax as long as the positions are tax efficient and subject to rates for qualified dividends. With nothing but a standard Boglehead portfolio I would say it is actually pretty difficult to get into the 24% bracket. Where things might get difficult is if you have a pension, real estate income, W2 income, consulting income, etc. If your taxable Social Security offsets your standard deduction, you would still have room for close to $200k of Roth conversions, so you would need to have perhaps $2M in Traditional IRA before I would be worried about hitting the 24% bracket.

Where you could get into trouble is if you were planning to be in the 10-12% brackets which is the territory you could get a bigger ACA subsidy by keeping your AGI low. Then you get hit with the double whammy of losing the subsidy and increasing your tax bracket.
"With nothing but a standard Boglehead portfolio I would say it is actually pretty difficult to get into the 24% bracket"
Agreed execpt for the specific topic we are focusing on here is "high net worth"
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

toddthebod wrote: Tue Jun 11, 2024 9:19 am
smitcat wrote: Tue Jun 11, 2024 8:55 am
TomatoTomahto wrote: Tue Jun 11, 2024 8:40 am
smitcat wrote: Tue Jun 11, 2024 7:32 am When folks are in this zone they are typically in a low tax bracket - even folks with 10"s of milions of $$$ in funds can easily be in the 24% bracket at that time which offers these opportunities.
Unless you have all 10s of millions of dollars in BRK, I think the 24% bracket is not “easily” reached. 2% is typical taxable income, then add in SS, any pensions, deferred compensation, etc.

We can agree to disagree and I’ll go bother someone else.

These are some paramaters to look at.
- $500K to no limit in Roth accounts
- a few million in municipal bonds
- $15 million or more in equity at say 1.5% Div average
- 401K up to $100K can be completely QCD's
- SS say $100K

$85K - SS taxable $85
$225K - 1.5% of 15 million
$310 total about $110 short of the 24% bracket.
A person could 'spend' $500K every year and never exceed the 24% bracket, a person could 'spend' a couple of million $$ in one year and never exceed the 24% bracket.
Your dividends mostly have no effect on your tax bracket.
Exactly - except by moving you thru 0% to 15% to 20% brackets for LTCG.
But we are focusing in HNW here which could have 'incomes' in retirement reaching the 24% bracket.
User avatar
FiveK
Posts: 16174
Joined: Sun Mar 16, 2014 2:43 pm

Re: Pre-tax vs Roth for HNW

Post by FiveK »

smitcat wrote: Tue Jun 11, 2024 7:38 am
FiveK wrote: Mon Jun 10, 2024 7:33 pm
smitcat wrote: Mon Jun 10, 2024 7:13 pm How do yo do this in the simplest form without using mutiple years for a comparison?
How do you configure the 'this year plan" vs 'future plan'?
1) Pick a representative year, such as "several years after starting RMDs" and calculate the marginal tax rate for that year using real returns, assumed withdrawal rate, and current tax law.
2) Determine the income for this year that will have this year's marginal tax rate match the one from step #1.
3) Repeat steps 1 & 2 next year.

Perfectly optimal? Probably not.
Close enough? Probably.

Perhaps the biggest "oh-by-the-way" is the extent to which "this year's" choice will affect the future. Given the ~$300K/yr range for MFJ filers that covers the 22% and 24% federal brackets, there is a lot of room for error for those with income in that range. :)
Correct me if I am wrong but this will only work if....
- the drawdown level is mostly keeping time with the rising RMD % (otherwise said - somwhere around a 4% draw)
- the persons portfolio AA's goals are to remain the same out to their demise (otherwose said - no large changes in AA over time)
- the person is not actively planning for heirs (charitable will show up in early years but heirs mostly show up at the end)
In reverse order:
- If planning for heirs - whether charitable or high-earning, but not both at the same time ;) - then use the heirs' marginal rate instead.
- a reasonable assumption, but in any case simply a special case of assuming some rate of return.
- as long as the marginal rate won't change significantly, it doesn't have to be constant. A few percent change over time is not significant in terms of the effect on quality of life, although it may change the purely optimum result.

One can be more precise with a complex model. Whether that ultimately leads to better accuracy is the open question, the answer to which becomes known only in hindsight. For those who want to explore the complexities, have at it. For those satisfied with a "close enough" approach, the complexity isn't needed for the overall strategy.

For "this year's" tactics, it does seem reasonable to use an accurate Tax estimation tool to avoid inadvertently going into one of the very high marginal rate zones/tiers the tax code contains, because then* one can be both accurate and precise.

*With the exception of IRMAA tiers that, for example, aren't known yet for 2025 yet will use 2023's income in 2025.
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

FiveK wrote: Tue Jun 11, 2024 3:17 pm
smitcat wrote: Tue Jun 11, 2024 7:38 am
FiveK wrote: Mon Jun 10, 2024 7:33 pm
smitcat wrote: Mon Jun 10, 2024 7:13 pm How do yo do this in the simplest form without using mutiple years for a comparison?
How do you configure the 'this year plan" vs 'future plan'?
1) Pick a representative year, such as "several years after starting RMDs" and calculate the marginal tax rate for that year using real returns, assumed withdrawal rate, and current tax law.
2) Determine the income for this year that will have this year's marginal tax rate match the one from step #1.
3) Repeat steps 1 & 2 next year.

Perfectly optimal? Probably not.
Close enough? Probably.

Perhaps the biggest "oh-by-the-way" is the extent to which "this year's" choice will affect the future. Given the ~$300K/yr range for MFJ filers that covers the 22% and 24% federal brackets, there is a lot of room for error for those with income in that range. :)
Correct me if I am wrong but this will only work if....
- the drawdown level is mostly keeping time with the rising RMD % (otherwise said - somwhere around a 4% draw)
- the persons portfolio AA's goals are to remain the same out to their demise (otherwose said - no large changes in AA over time)
- the person is not actively planning for heirs (charitable will show up in early years but heirs mostly show up at the end)
In reverse order:
- If planning for heirs - whether charitable or high-earning, but not both at the same time ;) - then use the heirs' marginal rate instead.
- a reasonable assumption, but in any case simply a special case of assuming some rate of return.
- as long as the marginal rate won't change significantly, it doesn't have to be constant. A few percent change over time is not significant in terms of the effect on quality of life, although it may change the purely optimum result.

One can be more precise with a complex model. Whether that ultimately leads to better accuracy is the open question, the answer to which becomes known only in hindsight. For those who want to explore the complexities, have at it. For those satisfied with a "close enough" approach, the complexity isn't needed for the overall strategy.

For "this year's" tactics, it does seem reasonable to use an accurate Tax estimation tool to avoid inadvertently going into one of the very high marginal rate zones/tiers the tax code contains, because then* one can be both accurate and precise.

*With the exception of IRMAA tiers that, for example, aren't known yet for 2025 yet will use 2023's income in 2025.
'
When using the simple approach we were stumped by these....

"In reverse order:
- If planning for heirs - whether charitable or high-earning, but not both at the same time ;) - then use the heirs' marginal rate instead."
It was easy to plug in the charitable goals, but did not plug in a future tax rate for heirs since the heirs would only represent a smaller portion of the total of the accounts compared prior to inheritance. How would/could that work out?

(the persons portfolio AA's goals are to remain the same out to their demise)
"- a reasonable assumption, but in any case simply a special case of assuming some rate of return."
- the persons portfolio AA's goals are to remain the same out to their demise (otherwose said - no large changes in AA over time)
A rising AA is affected by both the rate and the amount in that account. That became a rather complex calculation with the change in AA, change in returns and change in taxes calculated for the comparison.
-
(the drawdown level is mostly keeping time with the rising RMD %)
"- as long as the marginal rate won't change significantly, it doesn't have to be constant. A few percent change over time is not significant in terms of the effect on quality of life, although it may change the purely optimum result."
This is the one that really torpedoed the simple approach. As long as the spending is expending the accounts in a reasonable way the simple guess here works fairly well. But when the draw on the accounts after SS election goes down each 1% picking a future year to do a simple comparable is a problem.
User avatar
FiveK
Posts: 16174
Joined: Sun Mar 16, 2014 2:43 pm

Re: Pre-tax vs Roth for HNW

Post by FiveK »

smitcat wrote: Tue Jun 11, 2024 4:28 pm When using the simple approach we were stumped by these....
- If planning for heirs - whether charitable or high-earning, but not both at the same time ;) - then use the heirs' marginal rate instead."
It was easy to plug in the charitable goals, but did not plug in a future tax rate for heirs since the heirs would only represent a smaller portion of the total of the accounts compared prior to inheritance. How would/could that work out?
It's not possible to maximize both the amount left to charities (who incur a 0% tax) and high earning heirs (who incur a high tax rate). Pick your priority and go with it.
(the persons portfolio AA's goals are to remain the same out to their demise)
- a reasonable assumption, but in any case simply a special case of assuming some rate of return.
A rising AA is affected by both the rate and the amount in that account. That became a rather complex calculation with the change in AA, change in returns and change in taxes calculated for the comparison.
Yes, there can indeed be a wide range of possible results, with the $17K/yr vs. $151K/yr result just one example. Unfortunately we can't accommodate all those possibilities simultaneously. We take our best guess, maybe hedged however we choose, and go with it - one year at a time.
(the drawdown level is mostly keeping time with the rising RMD %)
- as long as the marginal rate won't change significantly, [the invested balance] doesn't have to be constant. A few percent change [in marginal rate] over time is not significant in terms of the effect on quality of life, although it may change the purely optimum result.
This is the one that really torpedoed the simple approach. As long as the spending is expending the accounts in a reasonable way the simple guess here works fairly well. But when the draw on the accounts after SS election goes down each 1% picking a future year to do a simple comparable is a problem.
Agreed that it might be a problem if aiming for the absolute optimum. For those now filing MFJ, not knowing when the survivor will start to file S is also a barrier to hitting that absolute optimum.

If one's personal objective function is "die leaving the largest possible inheritance" that might support different actions than if one's personal objective function is "being able to spend enough while alive to be happy". In either case, there is only one marginal tax rate a person can hit at the end of "this year".

As you and tibbitts noted in this Roth/Tax question post, having widely varying marginal tax rates from year to year is unlikely to be close to optimum. Picking a marginal rate target for "this year" that is likely to be maintainable for ~the rest of one's life has a good chance of being in the optimum ballpark. Or have you seen complex models that suggest otherwise?
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

FiveK wrote: Tue Jun 11, 2024 5:43 pm
smitcat wrote: Tue Jun 11, 2024 4:28 pm When using the simple approach we were stumped by these....
- If planning for heirs - whether charitable or high-earning, but not both at the same time ;) - then use the heirs' marginal rate instead."
It was easy to plug in the charitable goals, but did not plug in a future tax rate for heirs since the heirs would only represent a smaller portion of the total of the accounts compared prior to inheritance. How would/could that work out?
It's not possible to maximize both the amount left to charities (who incur a 0% tax) and high earning heirs (who incur a high tax rate). Pick your priority and go with it.
(the persons portfolio AA's goals are to remain the same out to their demise)
- a reasonable assumption, but in any case simply a special case of assuming some rate of return.
A rising AA is affected by both the rate and the amount in that account. That became a rather complex calculation with the change in AA, change in returns and change in taxes calculated for the comparison.
Yes, there can indeed be a wide range of possible results, with the $17K/yr vs. $151K/yr result just one example. Unfortunately we can't accommodate all those possibilities simultaneously. We take our best guess, maybe hedged however we choose, and go with it - one year at a time.
(the drawdown level is mostly keeping time with the rising RMD %)
- as long as the marginal rate won't change significantly, [the invested balance] doesn't have to be constant. A few percent change [in marginal rate] over time is not significant in terms of the effect on quality of life, although it may change the purely optimum result.
This is the one that really torpedoed the simple approach. As long as the spending is expending the accounts in a reasonable way the simple guess here works fairly well. But when the draw on the accounts after SS election goes down each 1% picking a future year to do a simple comparable is a problem.
Agreed that it might be a problem if aiming for the absolute optimum. For those now filing MFJ, not knowing when the survivor will start to file S is also a barrier to hitting that absolute optimum.

If one's personal objective function is "die leaving the largest possible inheritance" that might support different actions than if one's personal objective function is "being able to spend enough while alive to be happy". In either case, there is only one marginal tax rate a person can hit at the end of "this year".

As you and tibbitts noted in this Roth/Tax question post, having widely varying marginal tax rates from year to year is unlikely to be close to optimum. Picking a marginal rate target for "this year" that is likely to be maintainable for ~the rest of one's life has a good chance of being in the optimum ballpark. Or have you seen complex models that suggest otherwise?
"It's not possible to maximize both the amount left to charities (who incur a 0% tax) and high earning heirs (who incur a high tax rate). Pick your priority and go with it."
Lete me say this differently - I find it easy to plug in the charitable number for all years but then have the issue with inheritance which will be left at only the end year. A portion of the funds will be left for inheritance, most funds used for spending, some funds used for previously stated charity along the way.

"A rising AA is affected by both the rate and the amount in that account"...."Agreed that it might be a problem if aiming for the absolute optimum."
Picking the year for the test and getting a reasonable Roth guidance has been the problem when comparing any simple test.

“If one's personal objective function is "die leaving the largest possible inheritance" that might support different actions than if one's personal objective function is "being able to spend enough while alive to be happy". In either case, there is only one marginal tax rate a person can hit at the end of "this year".”
Our goals are not to leave the largest possible inheritance or to spend every dollar ourselves – it is a blend between the two as I would guess would be the case with many others. The simple techniques to pick a year and a rate based on this most common blended goal is a challenge.

“As you and tibbitts noted in this Roth/Tax question post, having widely varying marginal tax rates from year to year is unlikely to be close to optimum.”
Yes – engineering a 0% tax rate for a number of years after retirement was achievable by us and looked really attractive without looking at the entire picture( the entire lifetime of the 401K funds).

“Picking a marginal rate target for "this year" that is likely to be maintainable for ~the rest of one's life has a good chance of being in the optimum ballpark. Or have you seen complex models that suggest otherwise?”
The complex models allowed us to pick Roth conversion numbers which were optimized. We ran these possible future scenarios thru a variety of possibilities such as widely ranging portfolio performance and widely ranging draw rates.
We were also able to plug in our variable AA , charitable and view the resultant outputs for % Roth conversion amounts each run. In addition to measuring the value of the accounts at life’s end it was easy to assess the additional value (or not) of the converted amounts for heirs.

We really tried to get the simple tests to work out reasonably well but that approach became more complex and awkward as each of these issues came up. I believe the ability to pick a reasonably good comparison year to do a comparison becomes impossible due to things like: larger taxable account %, lower draw rate %’s, non-fixed AA’s, and considerations for heirs.

As I speak to friends about Roth conversions I am now sensitized to their situations and goals and would still recommend the simpler approach unless these other issues appeared to have a larger affect on their accounts and goals.
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

FiveK wrote: Tue Jun 11, 2024 5:43 pm
smitcat wrote: Tue Jun 11, 2024 4:28 pm When using the simple approach we were stumped by these....
- If planning for heirs - whether charitable or high-earning, but not both at the same time ;) - then use the heirs' marginal rate instead."
It was easy to plug in the charitable goals, but did not plug in a future tax rate for heirs since the heirs would only represent a smaller portion of the total of the accounts compared prior to inheritance. How would/could that work out?
It's not possible to maximize both the amount left to charities (who incur a 0% tax) and high earning heirs (who incur a high tax rate). Pick your priority and go with it.
(the persons portfolio AA's goals are to remain the same out to their demise)
- a reasonable assumption, but in any case simply a special case of assuming some rate of return.
A rising AA is affected by both the rate and the amount in that account. That became a rather complex calculation with the change in AA, change in returns and change in taxes calculated for the comparison.
Yes, there can indeed be a wide range of possible results, with the $17K/yr vs. $151K/yr result just one example. Unfortunately we can't accommodate all those possibilities simultaneously. We take our best guess, maybe hedged however we choose, and go with it - one year at a time.
(the drawdown level is mostly keeping time with the rising RMD %)
- as long as the marginal rate won't change significantly, [the invested balance] doesn't have to be constant. A few percent change [in marginal rate] over time is not significant in terms of the effect on quality of life, although it may change the purely optimum result.
This is the one that really torpedoed the simple approach. As long as the spending is expending the accounts in a reasonable way the simple guess here works fairly well. But when the draw on the accounts after SS election goes down each 1% picking a future year to do a simple comparable is a problem.
Agreed that it might be a problem if aiming for the absolute optimum. For those now filing MFJ, not knowing when the survivor will start to file S is also a barrier to hitting that absolute optimum.

If one's personal objective function is "die leaving the largest possible inheritance" that might support different actions than if one's personal objective function is "being able to spend enough while alive to be happy". In either case, there is only one marginal tax rate a person can hit at the end of "this year".

As you and tibbitts noted in this Roth/Tax question post, having widely varying marginal tax rates from year to year is unlikely to be close to optimum. Picking a marginal rate target for "this year" that is likely to be maintainable for ~the rest of one's life has a good chance of being in the optimum ballpark. Or have you seen complex models that suggest otherwise?
FiveK, a different response which is way overdue…..

An apology – my responses are often short and lack some details ….as well as descriptions and background. They may not be able to follow as easily as I intend and that I my fault completely as I am not a great nor fast typist and therefore a victim of my own deficiencies.

Thank you – over the years your knowledge on a host of subjects has helped up greatly. I have failed to thank you on the numerous occasions where your posts and/answers have helped. You have been very helpful on many subjects not just on these Roth conversions topics. Your choice to take the time to both help others and detail the methods are very appreciated. I want to thank you for all of your efforts in many of these posts.
The reasons we have tried to make our experience mostly parallel your posts is due to how much we respect and learn from your posts.
Thank you again for your participation on this board and for your specific help with so many posts I have lost count.
User avatar
FiveK
Posts: 16174
Joined: Sun Mar 16, 2014 2:43 pm

Re: Pre-tax vs Roth for HNW

Post by FiveK »

smitcat wrote: Wed Jun 12, 2024 7:06 am ...
Picking the year for the test and getting a reasonable Roth guidance has been the problem when comparing any simple test.
...
The simple techniques to pick a year and a rate based on this most common blended goal is a challenge.

“As you and tibbitts noted in this Roth/Tax question post, having widely varying marginal tax rates from year to year is unlikely to be close to optimum.”
Yes – engineering a 0% tax rate for a number of years after retirement was achievable by us and looked really attractive without looking at the entire picture( the entire lifetime of the 401K funds).

...The complex models allowed us to pick Roth conversion numbers which were optimized.
...
We really tried to get the simple tests to work out reasonably well but that approach became more complex and awkward as each of these issues came up. I believe the ability to pick a reasonably good comparison year to do a comparison becomes impossible due to things like: larger taxable account %, lower draw rate %’s, non-fixed AA’s, and considerations for heirs.
...
OK, so this is interesting: you have "Roth conversion numbers which were optimized" - this is a series of annual amounts, correct?

Given that, can you look backwards to determine
  • the ending marginal rate (on the last $1 if you don't go through any rate "spikes" or "mesas"; on the total dollars from the start of the last spike or mesa if you do) for each year. One would expect that to be similar if not identical from year to year.
  • If it is pretty much constant, what choice of "pick a year" would have led to the same result?
It's also possible that the "optimized" numbers won't settle to a relatively constant marginal rate from year to year. That's not because the true optimum would likely be so variable, but because
a) it could be a rather "flat" optimum, in which many different values provide similar end results, and/or
b) the tax code is so horribly nonlinear that numerical optimization algorithms struggle to cope.
User avatar
FiveK
Posts: 16174
Joined: Sun Mar 16, 2014 2:43 pm

Re: Pre-tax vs Roth for HNW

Post by FiveK »

smitcat wrote: Wed Jun 12, 2024 7:19 am FiveK, a different response which is way overdue…..

An apology – my responses are often short and lack some details ….as well as descriptions and background. They may not be able to follow as easily as I intend and that I my fault completely as I am not a great nor fast typist and therefore a victim of my own deficiencies.

Thank you – over the years your knowledge on a host of subjects has helped up greatly. I have failed to thank you on the numerous occasions where your posts and/answers have helped. You have been very helpful on many subjects not just on these Roth conversions topics. Your choice to take the time to both help others and detail the methods are very appreciated. I want to thank you for all of your efforts in many of these posts.
The reasons we have tried to make our experience mostly parallel your posts is due to how much we respect and learn from your posts.
Thank you again for your participation on this board and for your specific help with so many posts I have lost count.
You're welcome and thank you in return. These discussions are interesting and pleasant!

I do have great sympathy for the complex model approach, based on many years industrial experience with various numerical modeling techniques. It's also clear that such an approach is not to the liking of the general population, who merely want a "useful" model, and the simpler the better. ;)

After spending much time with a homegrown model that includes all relevant tax items and our specific cash flow situation, and experimenting with different gradient-based and direct search optimization routines, the result for us boils down to "convert into the 24% bracket each year, with 'how far into that bracket?' dependent on when we expect the survivor to start filing single and for how long." Thus I'm interested to see whether your results are qualitatively similar. If I recall correctly, picking any year 5-10 years (or so?) after RMDs start for our comparison year would have given us a result similar to the complex model's.
smitcat
Posts: 13560
Joined: Mon Nov 07, 2016 9:51 am

Re: Pre-tax vs Roth for HNW

Post by smitcat »

FiveK wrote: Wed Jun 12, 2024 11:14 am
smitcat wrote: Wed Jun 12, 2024 7:19 am FiveK, a different response which is way overdue…..

An apology – my responses are often short and lack some details ….as well as descriptions and background. They may not be able to follow as easily as I intend and that I my fault completely as I am not a great nor fast typist and therefore a victim of my own deficiencies.

Thank you – over the years your knowledge on a host of subjects has helped up greatly. I have failed to thank you on the numerous occasions where your posts and/answers have helped. You have been very helpful on many subjects not just on these Roth conversions topics. Your choice to take the time to both help others and detail the methods are very appreciated. I want to thank you for all of your efforts in many of these posts.
The reasons we have tried to make our experience mostly parallel your posts is due to how much we respect and learn from your posts.
Thank you again for your participation on this board and for your specific help with so many posts I have lost count.
You're welcome and thank you in return. These discussions are interesting and pleasant!

I do have great sympathy for the complex model approach, based on many years industrial experience with various numerical modeling techniques. It's also clear that such an approach is not to the liking of the general population, who merely want a "useful" model, and the simpler the better. ;)

After spending much time with a homegrown model that includes all relevant tax items and our specific cash flow situation, and experimenting with different gradient-based and direct search optimization routines, the result for us boils down to "convert into the 24% bracket each year, with 'how far into that bracket?' dependent on when we expect the survivor to start filing single and for how long." Thus I'm interested to see whether your results are qualitatively similar. If I recall correctly, picking any year 5-10 years (or so?) after RMDs start for our comparison year would have given us a result similar to the complex model's.
I am not sure exactly what you are asking or if I could supply an answer but allow me to summarize some of our experiences with Roth conversion choices these past 6-7 years.
Some of the basic things we have seen is that for persons looking at these Roth conversion options it is much more straight forward if they have all or most of these ingredients:
- Most funds in tax deferred
- Intention to ‘spend’ most of the funds during their lifetimes
- Goals of fixed AA’s for the most part
- Ages of couple more similar
- Low quantity of funds can only help
General overview
The ability to utilize a simpler method for a best fit determination is not too difficult with the above parameters. When we began to look at this the simpler methods for conversions it became much more complicated due to a host of issues all deviating from the above. About a year after we started looking into conversions but a year or so before we were able to be in a position to do our first conversions we began utilizing a more detailed calculator. That approach required more upfront time but once the setup was completed subsequent runs were made pretty quick. Being able to do ‘what ifs’ in a few minutes and get some data to review was a big jump if our ability to review the pros and cons of any variables. After a year or so using that single calculator, we also added another detailed calculator that had additional options for automatic calculations such as ‘auto Roth optimization” and “auto drawdown optimization” as well as others. These options allowed us to ‘toggle’ between manual plans and auto optimized plans which gave us some ideas we had not previously thought of. After using these two detailed calculators for years we have seen some interesting results, many of which we have acted on. For anyone reading this please remember that these results were for our numbers and our goals only ….and that everyone would need to run their own numbers and goals to see what future scenarios may be possible for them.
Some results
Please note that none of these numbers or %’s are exactly what we had calculated - but they are reasonably close. Also note that I have removed the charitable contributions from the 401K comparisons so that the total %’s would return to 100%.
- When we ran an early age of demise for one spouse it always preferred more conversions (more = higher % of total tax deferred converted to Roth)
- When we ran lower portfolio performance it mostly (but not always) favored lower conversions
- When we ran higher draw rates it mostly favored lower conversions
- The possibility of SS reduction per the SS site had a smaller affect
- The possibility of 2026 tax returns had a smaller affect
Once we had a bunch of feedback from the above runs we decided to make up our base case and run that with and without conversions as a base. Everyone has to make their own choices as to a base case but ours has some of these variables in there- both live long and to a similar age, portfolio performance mush lower than historical, draw rate that we had planned, tax rules as they stand now, etc. Using the detailed calculators they both recommended larger Roth conversions which made the first-year choice(s) rather easy. After reading more Bogle posts and after thinking about the results we had already seen last year we decided to run an array of variables that we were interested in and chart them for our review.
What we ran last year were two variables we wanted to see
- We ran four different portfolio performance numbers , about 25% of historical , about 50%, about 75% and 100%.
- We also ran four different draw numbers , about 75% of desired plan, 100% plan, 150% plan and 200%.
- We then had 16 results (4X4) plus the base case that we charted (this took about 1-1/2 to 2 hours total)
What we saw was this
- All of the runs recommended a muti year Roth conversion plan of at least 50% from where we were last year (not from initial tax deferred years back)
- Since we could not reasonably make that 50% Roth conversions in one year the 2023 conversion plan was now and easy choice or us
- Some of those runs recommended multiyear Roth conversions plans of near 100%
- The Roth conversion reflect a significant gain over non Roth conversions in this past year (as they had in the past years)
- In cases where we are curious about what ‘over conversions’ would ‘cost’ we just run that scenario (it is not that ‘costly’ in these test runs)
This next year we will not be running all of the same variables but we will be running a few newer variations because the more we look at these results and our updated situation the more we learn.
Also note that although all of these Roth conversion runs come out positive in our favor, we also need to adjust the ending results by the calculated benefit to our heirs.
Post Reply