Looking for the magic formula for Roth conversions vs RMDs

[Human]

I agree with @Scuba that one has to factor in the loss of growth on the dollars that would be used to pay taxes due to the conversion.

Anyway, I’ve spoken with CFPs and CPAs in the past about conversions. I’d go in excited and leave having learned that the factors are complicated, so much is unknown (what will the future tax laws be, for example), and/or conversations wouldn’t benefit me.

Still, every year I still consider a Roth conversion as part of my end of year process...and every year I kick the can down the road.

 

[growing_older]

Are the people that say this dumber than us, or smarter? Serious question. Of course all things being equal I'd rather defer; but if I'm pretty sure it's a choice of 15% of X now, or 25% of X (in constant dollars) later, the former is going to be better.

If you expect to have any IRA assets in your estate, there is an additional complication of tax rates for your heirs. Roth IRA assets are tax free to heirs but Trad IRA assets are taxed at the heir's tax rate, possibly requiring RMDs on their own accelerated schedule. If your heirs are in their peak earning years, this could be their highest tax rate.

 

[RunningBum]

You don't have to guess on the effect of growth and taxes, you can use math and assumptions to figure it out.


It's been years since I've done it but run a spreadsheet of $10K in a tIRA, and $10K in a Roth with taxes from the conversion paid from outside the Roth. Figure in a growth rate for both accounts, and also for the money from outside that would keep growing in the non-conversion case. Then figure taxes on liquidation of the tIRA, and on the growth in the outside account that you didn't have to use to pay taxes. There will be no taxes on liquidation of the Roth for the conversion case.


Assuming the conversion and liquidation tax rates are the same, or higher on the liquidation, and you use the same return on investment on the accounts, the Roth conversion will come out ahead. If the tax rate on the liquidation is lower, it's likely that you shouldn't convert.

 

[CardsFan]

You don't have to guess on the effect of growth and taxes, you can use math and assumptions to figure it out.

Isn't an assumption the same as a guess?

 

[pb4uski]

I was thinking about this and there is definitely time value of money to be considered. Paying a lot of taxes now to avoid potential taxes later may not pay off in the long run. Depends on the opportunity cost of how much return you’re giving up by paying taxes now that you otherwise wouldn’t have to pay for many years. ....

Not really.... it is differences in tax rates that is the more important determinant.

An example if the tax rate is the same with the time value of money. Let's say someone has $100k and is subject to a 15% tax rate now and later and the assumed investment earnings rate is 7% and time horizon is 10 year.

If they convert now, they start with $85k in their Roth after paying taxes and in 10 years that $85k grows to $167k.

If they don't convert now, their $100k tIRA grows in 10 years to $197k.... if they withdraw it an pay they tax they have $167k.

If you run the same scenario but starting with a $100k tIRA and $15k in after tax funds to pay the taxes, the convert scenario ends with a value of $197k and the not convert scenario ends with $194k... with the slight difference due not to the time value of money but due to taxes on the taxable account earnings (the $15k in taxable funds grows to $27k at the after-tax earnings rate of 5.95% [7%*(1-15%)].

 

[Cobra9777]

Not really.... it is differences in tax rates that is the more important determinant.

An example if the tax rate is the same with the time value of money. Let's say someone has $100k and is subject to a 15% tax rate now and later and the assumed investment earnings rate is 7% and time horizon is 10 year.

If they convert now, they start with $85k in their Roth after paying taxes and in 10 years that $85k grows to $167k.

If they don't convert now, their $100k tIRA grows in 10 years to $197k.... if they withdraw it an pay they tax they have $167k.

If you run the same scenario but starting with a $100k tIRA and $15k in after tax funds to pay the taxes, the convert scenario ends with a value of $197k and the not convert scenario ends with $194k... with the slight difference due not to the time value of money but due to taxes on the taxable account earnings (the $15k in taxable funds grows to $27k at the after-tax earnings rate of 5.95% [7%*(1-15%)].

+1

That's a good example. But conceptually, here's how I think about it... Embedded in your tax-deferred account is a tax liability that grows in lock-step with the taxable funds that could be used to pay tax in a conversion. IOW, when you continue to defer, the tax liability continues to grow by the same amount as your would-be tax dollars on a conversion. When you convert, it's really just like paying off a liability early in the hope that the liability will be less today than in the future (due to tax rates).

 

[redduck]

Would it be worthwhile for a 40-something who is still working and making $400,000 a year to start making these Roth IRA conversions? (We have a lot of these guys in the neighborhood).

 

[CRLLS]

Not really.... it is differences in tax rates that is the more important determinant.

An example if the tax rate is the same with the time value of money. Let's say someone has $100k and is subject to a 15% tax rate now and later and the assumed investment earnings rate is 7% and time horizon is 10 year.

If they convert now, they start with $85k in their Roth after paying taxes and in 10 years that $85k grows to $167k.

If they don't convert now, their $100k tIRA grows in 10 years to $197k.... if they withdraw it an pay they tax they have $167k.

If you run the same scenario but starting with a $100k tIRA and $15k in after tax funds to pay the taxes, the convert scenario ends with a value of $197k and the not convert scenario ends with $194k... with the slight difference due not to the time value of money but due to taxes on the taxable account earnings (the $15k in taxable funds grows to $27k at the after-tax earnings rate of 5.95% [7%*(1-15%)].

I got it. It makes sense. But now change the starting value from 100K to say 500 or 750K. And the starting age at, or near age 65. The conversion will possibly affect ACA subsidies, kick in IRMAA, and in some cases, Social Security tax torpedo. Doesn't that skew the benefit towards not converting?

For those who are planning on leaving a legacy, there may be reasons to make the conversion. But here is where I draw the line. 1st, our money is for DW and my benefit. Anything left over, and most likely, there will be a fair amount, whether it is taxable or not is a problem they will have to deal with.

 

[Huston55]

Not really.... it is differences in tax rates that is the more important determinant.

An example if the tax rate is the same with the time value of money. Let's say someone has $100k and is subject to a 15% tax rate now and later and the assumed investment earnings rate is 7% and time horizon is 10 year.

If they convert now, they start with $85k in their Roth after paying taxes and in 10 years that $85k grows to $167k.

If they don't convert now, their $100k tIRA grows in 10 years to $197k.... if they withdraw it an pay they tax they have $167k.

If you run the same scenario but starting with a $100k tIRA and $15k in after tax funds to pay the taxes, the convert scenario ends with a value of $197k and the not convert scenario ends with $194k... with the slight difference due not to the time value of money but due to taxes on the taxable account earnings (the $15k in taxable funds grows to $27k at the after-tax earnings rate of 5.95% [7%*(1-15%)].

You will find that Kitces says the same thing; more specifically, it’s “marginal” tax rates that make the difference. It took a while for me to get this into my head but, it is what it is.

https://www.kitces.com/blog/why-a-r...-idea-even-if-taxes-are-higher-in-the-future/

 

[misanman]

The general online procedure at Fidelity is:

- Create the Roth
- You will be asked how it is to be funded
- You will indicate a conversion of an existing Fidelity IRA
- You can also indicate whether or not it is a complete or partial conversion
- If a partial, then a list of securities are displayed and you can pick which ones are to be converted.
- Conversion is immediate

I would say the whole process is "easy peasy"

 

[RunningBum]

Isn't an assumption the same as a guess?

Yeah, I had that coming!

What I meant was, no reason to just wonder about you when you can actually do some math.

 

[RunningBum]

Would it be worthwhile for a 40-something who is still working and making $400,000 a year to start making these Roth IRA conversions? (We have a lot of these guys in the neighborhood).

Probably not, but we don't know the other side of the equation, what tax rate they will have in retirement. It's probably lower, so the answer would be No, but they could have a massive pension or large income from rentals or something like that.

 

[RunningBum]

I got it. It makes sense. But now change the starting value from 100K to say 500 or 750K. And the starting age at, or near age 65. The conversion will possibly affect ACA subsidies, kick in IRMAA, and in some cases, Social Security tax torpedo. Doesn't that skew the benefit towards not converting?

For those who are planning on leaving a legacy, there may be reasons to make the conversion. But here is where I draw the line. 1st, our money is for DW and my benefit. Anything left over, and most likely, there will be a fair amount, whether it is taxable or not is a problem they will have to deal with.

ACA subsidies could be a factor at any age before medicare, but if they are at 65, it's no longer a factor, right?

IRMAA, yes, a factor, but it will be a factor with MRDs too, right?

Likewise the SS tax torpedo. Around 65 you don't have to be taking SS, so maybe those 5 years are a good time to convert before you start adding SS income, if you have deferred.

The IRMAA and SS factors favor levelling out income over your remaining years, which would usually mean taking partial conversions, but perhaps leaving some for MRDs. I doubt that skipping conversion and taking larger MRDs will work better. There may be some individual factors, so it's worth working out the various ways with a spreadsheet or tax program.

 

[CardsFan]

Yeah, I had that coming!

What I meant was, no reason to just wonder about you when you can actually do some math.

Sorry, couldn't resist.

But you have a good point that you can get an idea of the benefit of conversions by running different scenarios. At least then you are making an educated guess.

 

[CRLLS]

ACA subsidies could be a factor at any age before medicare, but if they are at 65, it's no longer a factor, right?

IRMAA, yes, a factor, but it will be a factor with MRDs too, right?

Likewise the SS tax torpedo. Around 65 you don't have to be taking SS, so maybe those 5 years are a good time to convert before you start adding SS income, if you have deferred.

The IRMAA and SS factors favor levelling out income over your remaining years, which would usually mean taking partial conversions, but perhaps leaving some for MRDs. I doubt that skipping conversion and taking larger MRDs will work better. There may be some individual factors, so it's worth working out the various ways with a spreadsheet or tax program.

My gut feeling is that it greatly depends on the income/expense levels required for living. If you are close to ACA or IRMAA triggers with normal required income/expenses, then there is virtually no or little benefit to converting at all. It is an individual situation for sure. Even if one converts a small amount over several years, with a larger IRA/401K, the net benefit may be too small to really matter in the big picture.

In our picture, ACA subsidy was a consideration for the last few years. Now we are on Medicare, IRMAA is in play. I was bring this up in light of the fungible calcs that pb4uski presented. If one was near those triggers, then the tie goes to non-convert team. (I think).

 

[Happyras]

FWIW I was planning on doing conversions starting in 2018, (I am 63 and DW is 62), at our new marginal rate of 24% since we believe our marginal rate will be higher when RMD's kick in. Rates will likely go back up in 2026, and we will have significant RMD's if not converted pushing us into the next marginal bracket. However, they complicated this plan for us since we own a Sub S corp as well as rental income. This does not apply to many, but I thought I would share this complication as I know there are a few Sub S folks on this forum.

We can only take the 20% deduction for small business pass through income (QBI) up to an AGI of $315K (all sources) . Since we do not know for certain what our QBI will be until final books after year end, it makes it hard to know what amount to convert to stay below the 24% and the 315K limit for married filers. Most all of this so called QBI is retained earnings so the tax comes out of other income like pensions, and we actually see less net income to us than if we did not have a profit. This one factor alone will likely make it impossible to justify conversions, unless we make significant capital investment and Section 179 expense to eliminate QBI profits.

Making it worse, you can not re-characterize a conversion. Top this off with the new rule that makes all of our employee meal perks now subject to being full pass through income, the more pizzas they buy, the closer the unknown QBI approaches the combined AGI limit. (Not too happy about this one as we allow employees to buy lunches and evening meals when they are very busy. Obviously this is not a common issue. In the past we paid tax on 50% not the full amount.)

 

[pb4uski]

You will find that Kitces says the same thing; more specifically, it’s “marginal” tax rates that make the difference. It took a while for me to get this into my head but, it is what it is.

https://www.kitces.com/blog/why-a-r...-idea-even-if-taxes-are-higher-in-the-future/

Yes, I meant marginal tax rates but should have been more specific.

 

[pb4uski]

Would it be worthwhile for a 40-something who is still working and making $400,000 a year to start making these Roth IRA conversions? (We have a lot of these guys in the neighborhood).

Probably not, but we don't know the other side of the equation, what tax rate they will have in retirement. It's probably lower, so the answer would be No, but they could have a massive pension or large income from rentals or something like that.

+1 Only if they expect that their incremental tax rate on the conversion will be less than their marginal tax rate in retirement.

 

[pb4uski]

I got it. It makes sense. But now change the starting value from 100K to say 500 or 750K. And the starting age at, or near age 65. The conversion will possibly affect ACA subsidies, kick in IRMAA, and in some cases, Social Security tax torpedo. Doesn't that skew the benefit towards not converting?

For those who are planning on leaving a legacy, there may be reasons to make the conversion. But here is where I draw the line. 1st, our money is for DW and my benefit. Anything left over, and most likely, there will be a fair amount, whether it is taxable or not is a problem they will have to deal with.

It depends on what your other income is before the Roth conversions and what constraints exist. For example, if you were living off a huge slug of taxable savings in low-interest FDIC savings accounts then you might have some room for Roth conversions even if ACA subsidies is a constraint. There's probably no getting around the tax torpedo depending on your situation.

What I find best is start with constraint, then work backwards to the Roth conversion. The also test what happens if you do Roth conversions beyond the constraint and what the "net" economic cost of those conversions is.

 

[ILikeStarTrek]

Are you familiar with iORP? An acronym for Optimal Retirement Planner, this program seeks to optimize your retirement cashflows. One of the parameters it uses is Roth conversions to minimize taxes on RMDs. While not a perfect solution, it does provide some very useful data for consideration. Check it out at
https://www.i-orp.com/GOPtax/index.html

I tried running i-ORP but it didn't work for me, at least the Roth conversions part didn't work.

My initial inputs for Extended ORP were:
Current age 55
Tax-deferred initial balance 900
Roth initial balance 100
after-tax initial balance 200
Social security benefits 20
age to begin 67
Pension, not indexed to inflation 15
age to begin 67
Retirement age 55

Here's what it showed me for Federal Income tax bracket for first few years:
Age NoTax 10% 12% 22% 15% 25% Total Taxes Penalty Capgains
055 12 10 0 0 0 0 22 1 2 0
056 12 10 0 0 0 0 22 1 2 0
057 12 10 30 0 0 0 53 5 5 0
058 13 10 31 0 0 0 54 5 5 0
059 13 10 32 0 0 0 55 5 5 0
060 13 11 32 10 0 0 66 7 0 0
061 14 11 33 11 0 0 68 7 0 0

Then I went back and added
IRA to Roth IRA Conversions:
I tried picking choices ranging from 1 to 6. E.g. 6 is: 35% Tax bracket ceiling

I re-ran i-ORP with this addition. But it said:

Model M1811803JrQ4HzhqQ
will not solve, possibly because your Roth Conversion,
tax bracket limit may be over constricting your taxable income.

From age 55 to 59, my income tax bracket never even reaches the 22% bracket. Telling Roth conversions to go to the top of the 22%, 24%, or 35% brackets should be able to do some conversions during these years so it should not give me an error.

Does someone who has experience with this tool know why it generates that error when I try to add in Roth conversions?

 

[Scuba]

Would it be worthwhile for a 40-something who is still working and making $400,000 a year to start making these Roth IRA conversions? (We have a lot of these guys in the neighborhood).

Likely not, unless these folks expect to make more income after retiring.

 

[Scuba]

I tried running i-ORP but it didn't work for me, at least the Roth conversions part didn't work.



My initial inputs for Extended ORP were:

Current age 55

Tax-deferred initial balance 900

Roth initial balance 100

after-tax initial balance 200

Social security benefits 20

age to begin 67

Pension, not indexed to inflation 15

age to begin 67

Retirement age 55



Here's what it showed me for Federal Income tax bracket for first few years:

Age NoTax 10% 12% 22% 15% 25% Total Taxes Penalty Capgains

055 12 10 0 0 0 0 22 1 2 0

056 12 10 0 0 0 0 22 1 2 0

057 12 10 30 0 0 0 53 5 5 0

058 13 10 31 0 0 0 54 5 5 0

059 13 10 32 0 0 0 55 5 5 0

060 13 11 32 10 0 0 66 7 0 0

061 14 11 33 11 0 0 68 7 0 0



Then I went back and added

IRA to Roth IRA Conversions:

I tried picking choices ranging from 1 to 6. E.g. 6 is: 35% Tax bracket ceiling



I re-ran i-ORP with this addition. But it said:



Model M1811803JrQ4HzhqQ

will not solve, possibly because your Roth Conversion,

tax bracket limit may be over constricting your taxable income.



From age 55 to 59, my income tax bracket never even reaches the 22% bracket. Telling Roth conversions to go to the top of the 22%, 24%, or 35% brackets should be able to do some conversions during these years so it should not give me an error.



Does someone who has experience with this tool know why it generates that error when I try to add in Roth conversions?

I had a similar problem with i-Orp. You can email them and they will respond. In my case, I haven’t been able to spend more time on it yet. I’m going to try their suggestions to see if it will work for us as I feel we need a tool like this to help us plan. If I can’t get it to work, I’m going to get professional help to model our details.

 

[kramer]

You will find that Kitces says the same thing; more specifically, it’s “marginal” tax rates that make the difference. It took a while for me to get this into my head but, it is what it is.

https://www.kitces.com/blog/why-a-r...-idea-even-if-taxes-are-higher-in-the-future/

Thanks for linking this article .... it was well worth a read (dated 2014) and it turned out to be prescient with respect to tax reform this year.

To quickly summarize for those who haven't read it, there is a strong argument made by many that government will need more money in the future and thus tax rates will go up. However, there are lots of ways to raise tax revenue without raising marginal tax rates and some revenue increasing tax changes actually lower marginal tax rates. So one shouldn't automatically assume tax law changes will cause marginal tax rates to be higher in the future, even if the total tax burden is higher. (tax increases can be done by reducing deductions, introducing a VAT, expanding the base of people paying taxes or widening lower tax brackets, increased payroll taxes on earned income, etc)

 

[Curmudgeon]

I tried running i-ORP but it didn't work for me, at least the Roth conversions part didn't work.

I ran with your inputs and was able to reproduce your results. I think this is a recently-introduced bug; it looks as though the model was recently updated to reflect the new tax law with tax rate changes in 2018 and 2028. I would report this, the author is generally pretty responsive (not bad for a free tool!)

On the other hand, I chose option '7', unlimited Roth conversions, and it did indeed model such conversions, using up to the 22% bracket. Give that a try, and wait for the bug fix if you're uncomfortable with using 22%.

 

[Curmudgeon]

On the other hand, I chose option '7', unlimited Roth conversions, and it did indeed model such conversions, using up to the 22% bracket.

I thought, at first, that that model looked a bit odd, so I went back and looked at it some more, and I think it makes sense now. It is telling you to do about 5 years of IRA2ROTH conversions, going into the 22% bracket to do so. After that, your IRA balance becomes low enough that your RMDs will never cause you to exceed the 12%/15% bracket, and so no more conversion is desirable.