Deferral calculation on AGI, Taxable or marginal?

[Kings over Queens]

What you deferred at, or the effective rate of taxes avoided on deferrals is interesting but not really relevant to the Roth conversion decision. Typically, taxes paid on a Roth conversion will be a lower effective rate than the effective rate on the deferral.

I built out the spreadsheet because it IS interesting (#nerd) and because I needed to see the numbers to scratch that itch, good or bad. From 1993 to 2023 we deferred a total of $330,722 at an average effective rate of 13% The lowest year was $2,807 @ 1%, second lowest $15,674 deferred @ 6% and the highest was $10,104 @ 16%. Effective rate meaning tax paid divided by taxable income. That same time period comparing to AGI would be 9% rate.

I built out a separate page where I adjusted today's tax brackets to the year 2042 when RMD's kick in (birth year 1967) using 2.5% compounded inflation, the top of the 12% or bottom of the 22% bracket is $147,077.

Assume for a second I slow the growth in our deferral accounts to a 4% rate of return, today's $1.825 grows to $3.7 and spins off $150,289 in RMD's. This assumes no conversions or reductions due to plain income disbursements. Do I expect to earn 4% average over 18 years?

My best guess is those $150,000 in RMD's will or would be at somewhere between 11% to 13%.

My decision is made, at least for this year, we'll convert $85,000 to the bottom of the 22% bracket. Next year we expect a large dividend payment and will likely defer as much as possible to save on NIIT. 2026 is a wildcard. I should be retired and will not have any earned/wage income making room for consistent conversions, but will the tax brackets sunset?

First world problems.

 

[Gumby]

I have used a real rate of return, today's tax brackets and the marginal tax rate (including things like IRMAA and NIIT) to compare what added taxes I will pay when I take the money out by Roth conversions versus RMDs.

 

[Kings over Queens]

I have used a real rate of return, today's tax brackets and the marginal tax rate (including things like IRMAA and NIIT) to compare what added taxes I will pay when I take the money out by Roth conversions versus RMDs.

I don't know how you can't.

My only hesitation with the $85,000 conversion this year is that it will be done at a higher (17%) effective rate, which is both higher than when the deferral was made and my assumed/estimated rate at RMD.

This stuff can drive you batty.

I can wait until 2026 when I hope to be fully FIRED and won't have any wage/earned income clogging up the bucket. This assumes I don't die in a fiery crash or Skynet turns on us.

First world problems.

 

[pb4uski]

I built out the spreadsheet because it IS interesting (#nerd) and because I needed to see the numbers to scratch that itch, good or bad. From 1993 to 2023 we deferred a total of $330,722 at an average effective rate of 13% The lowest year was $2,807 @ 1%, second lowest $15,674 deferred @ 6% and the highest was $10,104 @ 16%. Effective rate meaning tax paid divided by taxable income. That same time period comparing to AGI would be 9% rate.

I built out a separate page where I adjusted today's tax brackets to the year 2042 when RMD's kick in (birth year 1967) using 2.5% compounded inflation, the top of the 12% or bottom of the 22% bracket is $147,077.

Assume for a second I slow the growth in our deferral accounts to a 4% rate of return, today's $1.825 grows to $3.7 and spins off $150,289 in RMD's. This assumes no conversions or reductions due to plain income disbursements. Do I expect to earn 4% average over 18 years?

My best guess is those $150,000 in RMD's will or would be at somewhere between 11% to 13%.

My decision is made, at least for this year, we'll convert $85,000 to the bottom of the 22% bracket. Next year we expect a large dividend payment and will likely defer as much as possible to save on NIIT. 2026 is a wildcard. I should be retired and will not have any earned/wage income making room for consistent conversions, but will the tax brackets sunset?

First world problems.

I think your analysis approach is flawed. You're looking at total effective tax rates... as you state, total tax paid divided by total income. I think that you want to lokk at the incremental tax rate and isolate tax-deferrals and the tax benefit of tax deferrals. Also, when looking at Roth conversions vs RMDs you want to isolate Roth conversions and taxes paid on Roth conversions compared to estimated RMDs and taxes paid on those RMDs.

Let's use an example of a single person in 2024 with $120,000 of earnings... would pay $18,338.50 of federal tax for a 15.28% effective rate.

Now if they deferred $20k into a 401k, they would have $100,000 of income and pay $13,841 in federal tax for a 13.84% effective rate.

By your calculations the benefit of deferral is negligible since it only reduces the effective tax rate from 15.28% to 13.84%.

But by deferring $20k, they save $4,497.50 in federal income tax or 22.49%. That is the relevant rate.

The first $4,875 of 401k deferrals saves 24% or $1,170 and the remaining $15,125 saves 22% or $3,327.50.

see IRS & State Tax Calculator | 2005 -- 2024

So for me as an example, I'll do ~$21k of Roth conversions in 2024 and pay an additional $2,520 in tax since we are in the 12% bracket before any Roth conversions and we convert to the top of the 12% tax bracket.

OTOH, if I don't do that $21k in Roth conversions then I'll end up with higher RMDs in the 12% and 22% tax brackets that blend to ~16% so I would rather pay 12% today than 16% in my mid 70s..

 

[Winemaker]

The genie is already out of the bottle. One must look at either converting at 22%, 24%, plus IRMAA, plus NIIT as a couple, or as a widow(er). We'll also get a clearer picture if the current tax rates are extended another 10 years.

 

[Kings over Queens]

I think your analysis approach is flawed. You're looking at total effective tax rates... as you state, total tax paid divided by total income. I think that you want to lokk at the incremental tax rate and isolate tax-deferrals and the tax benefit of tax deferrals. Also, when looking at Roth conversions vs RMDs you want to isolate Roth conversions and taxes paid on Roth conversions compared to estimated RMDs and taxes paid on those RMDs.

I don't disagree, the trouble I'm having if figuring out how much the difference would have been. I'm not expert level with excel and will try the calculator you provided. Thanks for that!

 

[pb4uski]

The difference between total effective rates and incremental effective rates would typically be substantial since incremental effective rates are usually a blend of your highest or second highest tax brackets (24% and 22% in the example used above) and total effective rates are a blend of 0% (income offset by deductions), 10%, 12%, 22% and 24% in the example above.

 

[Gumby]

The difference between total effective rates and incremental effective rates would typically be substantial since incremental effective rates are usually a blend of your highest or second highest tax brackets (24% and 22% in the example used above) and total effective rates are a blend of 0% (income offset by deductions), 10%, 12%, 22% and 24% in the example above.

Yes, total effective rate would be misleading, because you are converting or taking RMDs "on top of" your other income, so you want to look at the incremental increase..

 

[TheWizard]

Yes, total effective rate would be misleading, because you are converting or taking RMDs "on top of" your other income, so you want to look at the incremental increase..

We've kicked this around in previous threads.
My feeling is that retirement income that is more or less mandatory should be looked at being taxed at your EFFECTIVE, or average tax rate. This includes income such as SS, pensions, lifetime payout annuities, and RMDs. None of these income streams are truly "on top" of any others

Additional optional income, such as Roth conversions and perhaps withdrawals from tax-deferred prior to RMD age, should be analyzed at your MARGINAL tax rate...

 

[Gumby]

I do not find overall effective rate to be a useful metric for anything I want to know. But I concede that there are those who may want to know it for their own purposes. To your point, RMDs when required are "on top of" because you can avoid the income by using QCDs.

 

[RunningBum]

I do not find overall effective rate to be a useful metric for anything I want to know. But I concede that there are those who may want to know it for their own purposes. To your point, RMDs when required are "on top of" because you can avoid the income by using QCDs.

+1. Overall rate might be interesting info, but marginal rate is what makes sense for any optional income or deductions you are considering.