Roth Conversion Strategy Changes in 2018?

[Philliefan33]

Oh, I just thought of another variable. DH will start Medicare this year. I better check to see where we are in terms of having to pay more for Medicare Part B.

 

[kramer]

Oh, I just thought of another variable. DH will start Medicare this year. I better check to see where we are in terms of having to pay more for Medicare Part B.

good info here: https://www.kitces.com/blog/irmaa-m...-premium-surcharges-new-2018-magi-thresholds/

 

[Cobra9777]

In my personal situation, where taxable account is much larger than tax deferred, I would consider what you are proposing except that then all of my qualified dividends in my taxable account get taxed at 15% instead of 0%. So, my brackets for conversion go: 10%, 12%, 27%, 22%. For my situation, the 27% is just too high, so I am staying within the 12% bracket for the foreseeable future.

You are right, I had overlooked the 27% rate, but it doesn't really affect the conclusion very much. In our case, the taxable account is about half the size of tax-deferred. Our QDs subject to the 27% rate are relatively small. So if we converted to the top of the 22% bracket, the average rate paid (above the 12% bracket) would be 23%, and we could convert a much higher percentage of our tax-deferred balances.

That's still obviously higher than I would like to pay, but I have to compare to the rate we would otherwise pay at the time of RMDs + SS. If the rates revert, we'd pay 30% to the extent of QDs, then 25%, and possibly 28% or higher depending mainly on portfolio growth between now and then. In addition, as I mentioned before, I'm concerned that the cumulative effect of chained CPI will subject even more of the RMDs to the higher incremental rates, although I haven't tried to quantify that yet.

If the rates don't revert, we'd pay the same 23% average rate that we would have on the conversions, and possibly 24% or higher depending on portfolio growth. In that case, the rate advantage from converting beyond the 12% bracket is zero or very small. The main advantage would be tax-free growth over the next 10+ years, and also as protection against the scenario in which one of us dies and leaves the survivor paying single rates on huge RMDs.

 

[pb4uski]

This is one of my things to evaluate in 2018. I also am leaning towards more Roth conversions.... possibly to the top of the 22% tax bracket... to leave less in the tax-deferred IRA subject to RMDs once we start SS... and more growing tax-free quicker.

While the old 30%/new 27% marginal tax rate does turn my stomach, if I convert to the top of the 22% tax bracket, my effective tax rate on the entire converson would be ~18%... much more than the ~8-10% I pay now but probably much better than paying 25% or more later.

The extra tax on QD and LTCG is an issue but will be less so because our taxable funds are dwindling as we use that money for living expenses and to pay Roth conversion taxes.

Another issue for us is state income taxes... but those won't go away for a few years so for now I am like a deer in the headlights as to whether to convert to top of 12% or 22% but I have almost a year to figure it out.

 

[Philliefan33]

good info here: https://www.kitces.com/blog/irmaa-m...-premium-surcharges-new-2018-magi-thresholds/

Thanks for that link. The MFJ income where IRMAA kicks in is pretty high by our standards. I think we will be ok.

 

[RunningBum]

This is one of my things to evaluate in 2018. I also am leaning towards more Roth conversions.... possibly to the top of the 22% tax bracket... to leave less in the tax-deferred IRA subject to RMDs once we start SS... and more growing tax-free quicker.

Again, if you're going to 22%, why not 24%, unless 22% for a few years, then back to the top of 15% will be enough to empty your tIRA?

 

[pb4uski]

Interesting idea but I would only do that once we are clear of state income taxes in a few years... perhaps the best way to frame it might be to decide an optimal amount of RMDs once SS starts, then back into the resulting tIRA balance and then back into a conversion strategy to get to that balance by age 70.

 

[Cobra9777]

Again, if you're going to 22%, why not 24%, unless 22% for a few years, then back to the top of 15% will be enough to empty your tIRA?

Interesting idea but I would ony do that once we are clear of state income taxes in a few years... perhaps the best way to frame it might be to decide an optimal amount of RMDs once SS starts, then back into the resulting tIRA balance and then back into a conversion strategy to get to that balance by age 70.

Very interesting. I like the idea of getting more converted quicker to take advantage of tax-free growth longer. That helps reduce the risk of rate differentials going against you. I also like the idea of deciding on an acceptable tax-deferred balance at 70, and then working back from there. Lots of work to do...

 

[Sunset]

Very interesting. I like the idea of getting more converted quicker to take advantage of tax-free growth longer. That helps reduce the risk of rate differentials going against you. I also like the idea of deciding on an acceptable tax-deferred balance at 70, and then working back from there. Lots of work to do...

I'm not sure since both accounts would grow tax free (until taken out of IRA) that this has much positive effect, if one is paying a higher tax rate to do the conversion.
Especially if one needs to remove enough to pay the tax on the conversion from the IRA as well, or even sell stocks resulting in LTCG.

 

[RunningBum]

Interesting idea but I would only do that once we are clear of state income taxes in a few years... perhaps the best way to frame it might be to decide an optimal amount of RMDs once SS starts, then back into the resulting tIRA balance and then back into a conversion strategy to get to that balance by age 70.

My own optimal RMD amount would be 0, because with SS benefits, a small pension and dividends from my taxable account, any RMD is going to be taxed at 15+15%, and then 25%, (or 12+15%, then 22% if the new rates become permanent). I don't think I'd be able to take any RMDs at 12 or 15% without pushing dividends into being taxed. So I'd like to get my tIRA fully converted even if I have to do a lot of it at 22 and 24%.

Everyone's situation is different though. My own is complicated by the subsidy, which has become to big for me to skip, so I don't expect to be able to convert much at all in 2018.

 

[pb4uski]

It may not be so bad... if I take 85% of our SS and my pension less $24k standard deduction and back into the Roth needed to keep RMDs out of the 22% tax bracket, I will need to reduce my tIRA by 15-20% between now and 70 1/2.... so that would be 2-3% a year plus growth.

So it will be close but if we continue our path of converting to the top of the 12% tax bracket and I defer SS to 70 then we should be able to get close... what will help is as our taxable accounts dwindle with living expenses I'll be able to convert more each year for the next 4 years until DW starts her SS and then some for the next 4 years until I start my SS and RMDs begin. Also, the tIRA growth will moderate as it becomes more and more of our bond allocation... so between the two it will be close... we may end up in a higher tax bracket for a few years.

I can also do higher conversions once we are redomesticated to a no income tax state in a few years.

 

[Cobra9777]

I'm not sure since both accounts would grow tax free (until taken out of IRA) that this has much positive effect, if one is paying a higher tax rate to do the conversion...

Not sure I follow. We've got 13-14 years until 70, during which time the tax-advantaged portfolio will hopefully grow. Any of that growth that occurs in the tIRA *will* be taxed... either as an RMD or as a conversion prior to RMDs. Any of that growth that occurs in the Roth will never be taxed. So unless I'm missing something, converting sooner will allow more of the growth to escape taxation. This is a positive effect regardless of which way the rate differentials ultimately play out. Even converting at 22-24%, it should be positive or neutral. Negative is possible, but far less likely.

 

[pb4uski]

Initial analysis suggests that even if I continue to Roth convert to the top of the 12% tax bracket that I'll be in the 22% tax bracket once RMDs begin.... it will be close... I'll have to monitor year by year but it will be hard to avoid it... nice problem to have I guess.

 

[CRLLS]

Going thru my personal situation, I think I am finding the only significant reason for converting is if you can get the IRA and 401K's into a Roth for estate planning purposes. Admittedly we would have to pay taxes from those same IRA's and 401K's which reduces the amount actually put into the Roth. Every time I look at it, because we are 65/66 and our money is predominately in IRA's and 401K's, it will be close to a wash either way when considering it all tax brackets, IRMMA and beneficiary's after tax benefits.

Unless there is >5% benefit in the end (whenever that is) I consider things to be not worth the bother.

I am still evaluating though. And I can change at any time things change significantly.

 

[PatrickA5]

I, too am considering conversions up to the 22% bracket. Probably won't, but I might. Right now our tax deferred retirement savings are at 85% with Roth (Tax Free) at 15%. I'd love to even that out substantially over the next 10 years until RMD's. Ultimately, I'd like to have only enough in tax deferred to generate RMDs up to the standard deduction amount (currently $24K) - basically making those RMDs tax free (kind of). Using MFJ, we could have around $650K in tax deferred wherein RMD's would be below the standard deduction amount. Obviously, there is more to the puzzle such as taxation of SS, any tax cliffs that we might fall off, etc.
I always get my money's worth from buying TurboTax and playing with it doing these "what ifs".

 

[Sunset]

Initial analysis suggests that even if I continue to Roth convert to the top of the 12% tax bracket that I'll be in the 22% tax bracket once RMDs begin.... it will be close... I'll have to monitor year by year but it will be hard to avoid it... nice problem to have I guess.

While that initially is possible, another issue with RMD's is they increase each year in percentage taken out.

 

[pb4uski]

Yes, I'm seeing that in my spreadsheets. However, even though I'm in the 22% bracket slightly once RMDs begin, my overall federal tax rate is 3-11% of my total income and 9-15% of the Roth conversion or RMD, as the case may be... so definitely much less than the taxes I saved when I deferred that income which was solidly in the 25% bracket or higher.... so even though I am slightly in the 22% bracket.... I win.

Plus, once we move to a no income tax state I'll be saving even more since I was in a high income tax state when I deferred that income.

 

[audreyh1]

In my personal situation, where taxable account is much larger than tax deferred, I would consider what you are proposing except that then all of my qualified dividends in my taxable account get taxed at 15% instead of 0%. So, my brackets for conversion go: 10%, 12%, 27%, 22%. For my situation, the 27% is just too high, so I am staying within the 12% bracket for the foreseeable future.

I see that the 24% tax bracket ends at $315K for MFJ, whereas the 22% tax bracket ends at $165K. In between those two is the $250K threshold above which you pay an additional 3.8% Net Income Investment Tax on long-term cap gains and qualified dividends. Are people taking this into account when they calculate their cap gains effective tax brackets?

 

[Winemaker]

Has anyone used the Retiree Portfolio Model from Bogleheads?
The spreadsheet creates a Base case and a Full case models that you can compare various conversions, SS start times, etc. While it was just released using the new tax law, it pretty much showed similar results that I had with the older version, unless I am really screwing the numbers up. While DW and I collect pensions, have a modest rental income, and under 2.08% WR, the spreadsheet shows only $148,000 difference in portfolios.
We have 5 years before one of us reaches FRA, and our projected SS benefit at 70 will equal our current pension amounts. Currently, PA does not tax SS, pensions, or retirement account withdrawals. We are both in great health... it must be the red wine.

 

[kramer]

I see that the 24% tax bracket ends at $315K for MFJ, whereas the 22% tax bracket ends at $165K. In between those two is the $250K threshold above which you pay an additional 3.8% Net Income Investment Tax on long-term cap gains and qualified dividends. Are people taking this into account when they calculate their cap gains effective tax brackets?

Thanks for noting that ... it doesn't really affect me as a single as the 3.8% surcharge kicks in at $200K which is above the top of the 24% individual tax bracket which tops out at $157.5 ... however, as you noted, for a MFJ couple, a conversion to anywhere near the top of the 24% bracket would cause their capital gains and qualified dividends to be taxed at 23.8% federal. (and it may affect interest, regular dividends and short term capital gains, too, I am not sure).

 

[RunningBum]

I see that the 24% tax bracket ends at $315K for MFJ, whereas the 22% tax bracket ends at $165K. In between those two is the $250K threshold above which you pay an additional 3.8% Net Income Investment Tax on long-term cap gains and qualified dividends. Are people taking this into account when they calculate their cap gains effective tax brackets?

Good point, and another reason to model your situation with a tax program. I made a large contribution to a DAF last year, which gave me a lot of room for Roth conversions, but combined with a large LTCG I (barely) hit NIIT, so I stopped at the top of 0%, which was really 15% for LTCGs pushed into taxable, plus 3.8% NIIT on the last $253 converted. Anything conversions beyond that would've been taxed at 10+15+3.8% or 28.8%. I knew about NIIT before doing this but hadn't really thought about it applying to me until I realized it did.

 

[kramer]

For a pretty easy to use spreadsheet to calculate your potential marginal rates at tax bomb time, select the Marginal Tax Rates spreadsheet to see a description and download:

Bob Hinkley Downloads

This site is from #Cruncher from the bogleheads site, he provides super helpful information on the forum. He has updated to include 2018 tax information (and also retains 2017, etc).

NOTES:
* In order for the Excel spreadsheet to load properly in Libre Office, I had to update to the most recent version, 5.4.4.2, even though I had a relatively recent version already.
* One thing that I do in my modeling is I adjust the Social Security tax threshold (for the 50% and 85% figures) by assuming it won't be adjusted for inflation, which it is not currently. For me, looking 17 years in the future, I multiply these numbers by 0.7, so instead of 25,000 and 34,000, I changed them to 17,500 and 23,800. You need to unprotect the Tables spreadsheet to do this, you can do that from the Tools menu when you are on the Tables spreadsheet and then modify the numbers.

 

[jroon]

Thanks for all the great responses! I learned quite a bit from this thread and the Boglehead thread. There is a lot to consider.

I am thinking of converting to the top of the 12% bracket this month; if the market tanks significantly during the year, I may convert another 50K or so.

 

[Trooper]

I am thinking of converting to the top of the 12% bracket this month; if the market tanks significantly during the year, I may convert another 50K or so.

Just curious - what does market performance have to do with conversion amounts? Taxable income is based on how much is converted at the time of conversion, not ending balance, correct?

 

[user5027]

I am thinking of converting to the top of the 12% bracket this month; if the market tanks significantly during the year, I may convert another 50K or so.

Just curious - what does market performance have to do with conversion amounts? Taxable income is based on how much is converted at the time of conversion, not ending balance, correct?

Correct.

jroon is considering a market timing strategy by converting a large number of shares when they are "cheap" and anticipating they will rebound.