Should We Make More Roth Conversions?

[JDSwim3]

Background Data:
Retired
Both age 61 (RMDs start at 72)
Including Pension and SS (her at 62, me at FRA), living expense withdrawal from tIRA will average 2.5% through RMD start (assumed 5% annual ROI, 3% historic annual inflation going forward)
Balance in tIRA - $1.57M (note - Roth balance is $355K all in SWTSX)
o SWAGX - $588K
o SWTSX - $477K
o SWISX - $271K
o VTSAX - $92K
o SCHP - $139K
EF / Market Slump Hedge (HYSA and I-Bonds) - $118K
Cash for additional prospective conversions in taxable brokerage - $110K

If ROI ticks up to even 7%, we would almost certainly find ourselves in a higher tax bracket come RMD time, which classically would say to make further conversions to slow tIRA growth.

Just want to make sure I'm accounting for all the variables (IRMAA, NIIT, RMD, TCJA expiration, etc...) in making a decision to either forego further conversions, go to top of 22%, or even 24%. As shown above, we have the cash for further conversions (or even a massive one this year to 24%), but I just get a bad feeling forking over that much to Uncle Sam in one fell swoop. Maybe just take it to top of 22% for two more years...?

Appreciate the collective thoughts of this intelligent readership!

[Deontic]

RMDs start at age 73 for you.

[retiredjg]

What about somewhere in the 24% bracket (not necessarily the top) for 2023 and 2024 and then just up to IRMAA after that (assuming you turn 63 in 2025).

This should trim but not eliminate the tax-deferred account before RMDs start.

[jeffyscott]

With similar assets and ages, I am only going to the top of the 12% bracket. But perhaps your pensions and SS are much larger than ours? I'm planning to shut down any conversions after one more year as we will have to rely on ACA for my spouse's coverage for a few years and the effective Federal marginal tax rate on conversion will be about 28% for us, due to the structure of ACA subsidies.

We expect to remain in the 12% bracket until there is a sole survivor and heirs will likely also be in the same.

If returns are high and it retroactively turns out it would have been advantageous to convert at 22%, the penalty is small since the next bracket is only 24%. But then if returns are high, we have even more money than expected so paying 2 points more in taxes is of little consequence. (After several years of conversions, we now have about 50% in Roth accounts and nearly all (90% ) of assets in TIRAs are bonds/fixed income. So there's little chance of high returns there, anyway)

OTOH, if returns are low and we've converted at 22%, there's a chance that we paid 22% on assets that could've come out at 12%.

I'm highly unlikely to suggest anyone convert even at 22%*, but for those who have so much that they're hitting whatever bracket comes after 24%, see #4 here:
viewtopic.php?p=7382677#p7382677

*I generally skip posts that mention things like being in tax brackets that start with a 3 or annual income/spending of hundreds of thousands of dollars.

[jeffyscott]

RMDs start at age 73 for you.

Actually it's 75 for anyone born in 1960 or later, which would include the OP and spouse.

[theblaze]

[Link fixed by admin LadyGeek (was https://ibb.co/3pdsxqH)]

[RetiredAL]

I am 73 and already doing RMD's. I am currently looking as to whether to re-start conversions or not.

Under our current MFJ filing, there is no need to more conversions. However, when "single" happens to the survivor, the convoluted SS taxation formulas in force today, will result in $10K to $15K of the needed spending withdrawal into a 40% SS hump before returning to 22%. While still filing MFJ, any conversion $ (above the RMD) will only be taxed at 22%. Removing another $100K from my IRA balances, get the survivor away from that 40% hump.

The calculations and decision points for doing this are very specific to the individual. For a higher level view, consult the BH Wiki on SS taxation -- "SS Heat Tables" for both MFJ and Single.
https://www.bogleheads.org/wiki/Taxatio ... y_benefits

[FiveK]

I am 73 and already doing RMD's. I am currently looking as to whether to re-start conversions or not.

Under our current MFJ filing, there is no need to more conversions. However, when "single" happens to the survivor, the convoluted SS taxation formulas in force today, will result in $10K to $15K of the needed spending withdrawal into a 40% SS hump before returning to 22%. While still filing MFJ, any conversion $ (above the RMD) will only be taxed at 22%. Removing another $100K from my IRA balances, get the survivor away from that 40% hump.

Paying 22% that avoids 40% is indeed a good strategy. How To Calculate The Marginal Tax Rate Of A Roth Conversion goes into more details.

The calculations and decision points for doing this are very specific to the individual. For a higher level view, consult the BH Wiki on SS taxation -- "SS Heat Tables" for both MFJ and Single.

The tables are good when it's only SS and ordinary income. Throw in qualified dividends and/or long term capital gains and the marginal rates can be even higher. See the second chart under Married taxpayers for one example.

[FiveK]

Just want to make sure I'm accounting for all the variables (IRMAA, NIIT, RMD, TCJA expiration, etc...) in making a decision to either forego further conversions, go to top of 22%, or even 24%. As shown above, we have the cash for further conversions (or even a massive one this year to 24%), but I just get a bad feeling forking over that much to Uncle Sam in one fell swoop. Maybe just take it to top of 22% for two more years...?

Much depends on what you expect for the marginal rate on future traditional withdrawals. At one extreme there is "all Qualified Charitable Distributions while alive and charitable organization heirs" with a future marginal rate of 0%: "no Roth conversions at all" as the logical consequence. The other extreme is something like "many years of the survivor filing single and heirs likely to be hit with 30+% marginal rates on forced distributions: Roth conversions into (maybe all the way through) the 24% bracket as the logical consequence.

Roth Conversion with Social Security and Medicare IRMAA describes a way to account for all the variables you listed, except for what will happen upon TCJA expiration. If I had to bet, I'd bet on "something other" than either "maintain the current bracket and rate structure" or "revert to the pre-TCJA bracket and rate structure" but exactly what that would be I have no idea.

[RetiredAL]

I am 73 and already doing RMD's. I am currently looking as to whether to re-start conversions or not.

Under our current MFJ filing, there is no need to more conversions. However, when "single" happens to the survivor, the convoluted SS taxation formulas in force today, will result in $10K to $15K of the needed spending withdrawal into a 40% SS hump before returning to 22%. While still filing MFJ, any conversion $ (above the RMD) will only be taxed at 22%. Removing another $100K from my IRA balances, get the survivor away from that 40% hump.

Paying 22% that avoids 40% is indeed a good strategy. How To Calculate The Marginal Tax Rate Of A Roth Conversion goes into more details.

The calculations and decision points for doing this are very specific to the individual. For a higher level view, consult the BH Wiki on SS taxation -- "SS Heat Tables" for both MFJ and Single.

The tables are good when it's only SS and ordinary income. Throw in qualified dividends and/or long term capital gains and the marginal rates can be even higher. See the second chart under Married taxpayers for one example.

In previous years, those have not been a concern as our taxable was near zero. That is changing some this year and will a lot next year, so those will become a concern to me.

[cas]

Background Data:
Retired
Both age 61

[. . .]

Just want to make sure I'm accounting for all the variables (IRMAA, NIIT, RMD, TCJA expiration, etc...) in making a decision to either forego further conversions, go to top of 22%, or even 24%.

I'm not seeing that you mention how you are getting health insurance prior to Medicare at age 65. (And I also looked back at your previous posts, and people asked about ACA participation, but I didn't see an answer.)

If you are getting health insurance via the ACA, then "ACA credit phase-outs" is another big variable you need to add to your IRMAA, NIIT, RMD, etc. list.

For more information: Roth Conversion and Capital Gains On ACA Health Insurance

[Parkinglotracer]

RMDs start at age 73 for you.

Actually it's 75 for anyone born in 1960 or later, which would include the OP and spouse.

Indeed 75. I am born in 1961 and am in the 75 group

[JDSwim3]

I'm not seeing that you mention how you are getting health insurance prior to Medicare at age 65. (And I also looked back at your previous posts, and people asked about ACA participation, but I didn't see an answer.)

I should have mentioned in my background data that my retirement includes pretty solid medical coverage with minimal premiums and co-pays until Medicare age. So ACA is a non-factor.

[smitcat]

I'm not seeing that you mention how you are getting health insurance prior to Medicare at age 65. (And I also looked back at your previous posts, and people asked about ACA participation, but I didn't see an answer.)

I should have mentioned in my background data that my retirement includes pretty solid medical coverage with minimal premiums and co-pays until Medicare age. So ACA is a non-factor.

Summarized well by FiveK above your uses for the funds are a consideration since your draw seems very low from what you posted.
Dependent upon your goals for charity or heirs (and heirs tax rates) will directly affect your future models on the potential value of conversions.
We utilize a calculator for these purposes and run scenarios with various portfolio performance numbers and various draw rates.

[Jack FFR1846]

Having medical covered until Medicare time, I'd be looking at your income at 73, RMD time. With just a little gain, I think you'll be looking at first year requirement of around $60k. (you could certainly do a more accurate calculation). I would look at what Medicare considers for income against their IRMAA extra cost adder for part B and part D. They take income from 2 years before you start Medicare but you can submit an application to ask for Recalculation if your income is less in your first year of Medicare. I did that. My income from 2 years ago was 2 clicks up the IRMAA ladder, roughly doubling my premiums. I have applied but Medicare is as slow as social security is.

I keep a keen eye on what the max income is (W2, dividends on taxable, interest, 1099, Roth conversions) and keep it all on a spread sheet and make sure I'm under that limit. This year, to give you an idea, the first point where they charge more is $194k for married filing jointly. I just retired mid year so did Roth conversions, giving myself some room, just in case I miscalculated ESPP, RSU or dividends are more than expected. I am also converting interest generating cash into investments that don't pay dividends to reduce this for next year.

Another thing to keep in mind is that in 2026, tax rates revert, adding about 2% per bracket.

I know there are very complicated tools to optimize Roth conversions. I did a very simple check, finding that I would have to convert over $300k a year to get my tIRA balance down near zero. Not doing that, so I'll just convert up to the IRMAA limit going forward.

[JDSwim3]

Summarized well by FiveK above your uses for the funds are a consideration since your draw seems very low from what you posted.

Don't anticipate QCDs unless ROI goes through the roof. Our seemingly low draw is based on a reasonably comfortable budget developed from several years of tracking every expenditure across many categories (maybe just this side of obsessively tracking). That tracking enabled establishment of our retirement budget and allowed me to retire fairly early, which is working so far 3 years in.

I built an Excel portfolio longevity model with key variable inputs [ROI, inflation, SS start, husband kick the bucket year (SS reduction, pension reduction, and single filer tax implications for her)] to validate and tweak the annual budget at sustainable draw levels, as needed, based on portfolio balance each January 1.

Our LTC will be self-funded, so our annual budgets are designed to provide adequate portfolio residual down the road. If LTC isn't needed, then Momma can splurge and the heirs get some windfall.

[jeffyscott]

[Link fixed by admin LadyGeek (was https://ibb.co/3pdsxqH)]

But the IRS is only showing what is pertinent right now, not what the rules will be in 2033 and beyond when it moves to 75.

https://www.google.com/search?q=rmd+starting+at+age+75

[jeffyscott]

Under our current MFJ filing, there is no need to more conversions. However, when "single" happens to the survivor, the convoluted SS taxation formulas in force today, will result in $10K to $15K of the needed spending withdrawal into a 40% SS hump before returning to 22%. While still filing MFJ, any conversion $ (above the RMD) will only be taxed at 22%. Removing another $100K from my IRA balances, get the survivor away from that 40% hump.

That makes sense, and the impact of the SS taxation thresholds not being indexed to inflation, will effectively move the hump to lower real income levels over time.

The calculations and decision points for doing this are very specific to the individual.

As and example of that, we have pensions, that by the time we actually are subject to RMDs, will likely have increased in nominal terms and put a potential sole survivor well beyond the hump. So there's no possible avoidance strategy in our case.

[heyyou]

Should We Make More Roth Conversions?

Of course it depends, but having done the Roth conversions for more than a decade after early retirement in 2005, these days the growth in my un-taxed, all-stock RIRA is quite welcome in my RMD period.

The only thing worse than paying taxes early is having to pay even more tax later. Calculate what fits your situation.

[Boglenaut]

Just want to make sure I'm accounting for all the variables (IRMAA, NIIT, RMD, TCJA expiration, etc...) in making a decision to either forego further conversions, go to top of 22%, or even 24%. As shown above, we have the cash for further conversions (or even a massive one this year to 24%), but I just get a bad feeling forking over that much to Uncle Sam in one fell swoop. Maybe just take it to top of 22% for two more years...?

Appreciate the collective thoughts of this intelligent readership!

JDSwim3,

I have been researching this topic like crazy this year. What I learned is that you need to map it out year-by-year with your own factors. Rules-of-thumb or generalization don't work very well here. Plan a full day or two working on a spreadsheet if you have such skills. It can be extremely enlightening.

Here is the list of factors I made to consider for us:

Probability one of you may be a single filer someday. Unless you can successfully plan to die the same year, this will happen.
IRMAA
ACA possibility
2026+ brackets/standard deduction/rates
Investment returns
If planning to leave to heirs, heirs future bracket
NIIT 3.8% on dividends, .9% payroll
How will pay for conversions? (we do from payroll while working)
SS Tax Rates
How long will each spouse work
ROTH IRA/401K eligibility/availability
Future traditional IRA/401 contributions
Pensions (she will have a tiny one)
Will you make non-conversion non-RMD withdrawals from traditional IRA/401ks
Capital gains tax brackets
Chance of changing states to higher/lower state income tax
My state is thinking about reducing income tax, will they?
Probability of losing college tax credits/benefits due to income, I Bonds


I don't know your Schwab tickers, but make sure to leave higher growth funds in Roth.

For us, I decided our traditional IRA/401Ks were right around $1M, so we will do net 100K conversions in 2023, 2024, and 2025 (to the top of the 24% bracket). After that, we will play it by ear because the case won't be as strong then. My wife gets 12% traditional "match" in her 401k, so it will go up even after we convert. Even if I am wrong and over-convert, one can think of this as a risk-reduction strategy where we protect against investments doing very well, unplanned tax increases, new income based surcharges,etc.

The best you can do is make an educated guess... there is no way to know the perfect solution. Reasonable minds can and will cometo different conclusions because there are so many unknowns.

[Boglenaut]

I am 73 and already doing RMD's. I am currently looking as to whether to re-start conversions or not.

Just make sure to do the RMD for the year BEFORE doing the conversions. You probably know this already, but I saw a few threads here where people missed this and made a mess.

[placeholder]

With similar assets and ages, I am only going to the top of the 12% bracket.

The 12% bracket is tricky if you hold taxable stocks with qualified dividends because the bracket is already at least partially filled with dividends at 0% so as you add ordinary income it eventually bumps out dividends from the 0% to the 15% bracket giving a net federal rate of 27%.

[jeffyscott]

With similar assets and ages, I am only going to the top of the 12% bracket.

The 12% bracket is tricky if you hold taxable stocks with qualified dividends because the bracket is already at least partially filled with dividends at 0% so as you add ordinary income it eventually bumps out dividends from the 0% to the 15% bracket giving a net federal rate of 27%.

Yes, technically I would only take total income up to the top of the 0% cap gain bracket, but the difference is only $200.

IOW, I will convert such total taxable income, including any dividends or cap gains, is less than $89,250 this year.

[RetiredAL]

I am 73 and already doing RMD's. I am currently looking as to whether to re-start conversions or not.

Just make sure to do the RMD for the year BEFORE doing the conversions. You probably know this already, but I saw a few threads here where people missed this and made a mess.

Yes I am aware, but thanks for bringing it up for other readers that may not know of this key sequencing detail.

My last periodic IRA withdrawal to our spending account is Nov 5, all of the QCD's have already processed, so all I will have left to do before a conversion is a manual lump withdrawal of $4100 to meet the RMD requirement. Additionally, and I don't know if this is an interaction concern or not, but my Inherited IRA has already had it's RMD accomplished via a QCD.

I will also have achieved 'safe harbor' tax status sometime in Oct via our SS with-holdings.

[smitcat]

Summarized well by FiveK above your uses for the funds are a consideration since your draw seems very low from what you posted.

Don't anticipate QCDs unless ROI goes through the roof. Our seemingly low draw is based on a reasonably comfortable budget developed from several years of tracking every expenditure across many categories (maybe just this side of obsessively tracking). That tracking enabled establishment of our retirement budget and allowed me to retire fairly early, which is working so far 3 years in.

I built an Excel portfolio longevity model with key variable inputs [ROI, inflation, SS start, husband kick the bucket year (SS reduction, pension reduction, and single filer tax implications for her)] to validate and tweak the annual budget at sustainable draw levels, as needed, based on portfolio balance each January 1.

Our LTC will be self-funded, so our annual budgets are designed to provide adequate portfolio residual down the road. If LTC isn't needed, then Momma can splurge and the heirs get some windfall.

We had also tracked expenses in great detail for many years due to treating our budget just like a small business. With your low draw and no planned QCD there may be a compelling reason for converting - I would (and did) run various most likely scenarios in a program like 'Pralana' to see what level of % Roth optimizes each scenario. This was very educational for us because we did not expect all of the results which we ended up with.
We previously had a homemade model which was handy but did not reflect all inputs/outputs nor was it 'tested' by enough users for confirmation of accuracy. With heirs as the end destination for some of the portfolio you need to make sure you adjust the ending account balances for all the tax affects at that transition.