Further thoughts on converting a traditional IRA to a Roth.

[Ed_The_Gypsy]

With the Tax Torpedo in mind, I am looking for a conversion strategy. Using a relatively simple spreadsheet and on-line tax estimators to compare different cases, and after wringing out the obvious bugs, I have learned some things that may be helpful to others, even though it may not be possible to generalize from our situation. This close to 70.5, our options are limited.

Basis:
1) We are 66. There is a reasonable possibility that I will ‘retire’ in Jan ’14 at 67, hence this homework.
2) DW is taking SS now. I have filed-and-suspended at my FRA this year and plan to take my SS at 70-1/4, in Jan of 2018, the year I turn 70.5 and MRDs begin. We would have a 4-year window to convert while our total SS is solely at my wife’s lower amount. I considered taking my SS at 67, but it reduces the survivor’s benefits for my wife, so this is out of the question. We will get a sizable combined SS eventually.
3) Virtually all of our assets except the house are in a modest traditional IRA. We each will get a one-time exemption for capital gains when we sell the house (and the condo when we go to the Home), so that should have no negative impact on these calculations. We own no other real estate.
4) Projections were made for both of us to age 100, just for grins.
5) An after-tax income starting at 67 from all sources was assumed, based on FIREcalc portfolio survivability plus some. In the first year, this is equivalent to over 8% from the IRA (for taxes and spending money). This high drain will only be for the first 4 to 7 years and is unavoidable unless I want to take SS before 70. (It is only about 1/3 of what my normal income used to be in North America, by the way.)
6) SS and IRA distributions were assumed to be our only income. Taxes will be paid from the IRA distributions.
7) After-tax income grows uniformly at an assumed 3%, SS payments at 2%, ROI is assumed to be 6%.

Cases:
A) The minimum case. The IRA does not get emptied.
From 67 to 71, withdraw from the IRA only enough to give us the target after-tax income and pay the taxes.
The IRA will not be empty by 71, so take RMDs from 71 onwards.
I take SS at 70-1/4.
B) Again, the IRA does not get emptied.
From 67 to 71, withdraw from the IRA up to the top of the 15% marginal tax bracket. Pay the taxes out of that, give us the target after-tax income and convert what is left over to the Roth.
The IRA will not be empty by 71, so take RMDs from 71 onwards.
I take SS at 70-1/4.
C) Empty the IRA.
From 67 to 71, withdraw from the IRA up to the top of the 15% marginal tax bracket. Pay the taxes out of that, give us the target after-tax income and convert what is left over to the Roth.
I take SS at 70-1/4.
At 70.5, we will be in the 25% marginal bracket (the Tax Torpedo). Withdraw from the Ira to the top of the 25% bracket. Pay the taxes out of that, give us the target after-tax income and convert what is left over to the Roth. The IRA will be empty in 3 years.
D) Empty the IRA faster.
From 67 to 71, withdraw from the IRA up to the top of the 25% marginal tax bracket. Pay the taxes out of that, give us the target after-tax income and convert what is left over to the Roth.
The IRA will be empty after those four years, before the MRDs start at 70.5.
I take SS at 70-1/4.

Summary of each case:
A) 67-71: minimum conversion at 15%. IRA still lives.
B) 67-71: maximum conversion to top of 15%. IRA still lives.
C) 67-71: maximum conversion to top of 15%.
71-73: convert to top of 25% each year. Empty IRA by 74.
D) 67-71: maximum conversion to top of 25%. Empty IRA by 70.5.

At age 100, the same total after-tax income in all cases. Results at age 100 by case:
A) Total taxes paid = $128k. IRA still has $18k. Roth has $473k left. (Total = $491k.)
B) Total taxes paid = $77k. IRA still has $12k. Roth has $560k left. (Total = $572k.)
C) Total taxes paid = $114k. IRA zero at 74. Roth has $288k left.
D) Total taxes paid = $105k. IRA zero at 70.5. Roth has $253k left.

Lessons learned:
1) Before 70.5, anything you can convert at 15% is good.
2) Converting the max at 15% before 70.5 has the best overall outcome: least taxes and maximum residual assets.
3) To empty the IRA, I pay more than the minimum taxes. This simplifies my future taxes because at some early point I will have no taxable income. I may not have to file quarterly estimates of income or file annually at all. This would be nice in my dotage—to be totally free of the Feds!
4) Interestingly, emptying it fast in 4 years at the 25% rate costs less tax than emptying slow (15% for 4 years, then at 25% for 3 years), because it gets emptied before I start my SS. However, I have less in the Roth all the way to the end, because I start drawing down the Roth 3 years earlier in Case D than Case C.
5) As a sub-case of Case C, I did consider postponing my taking SS until after the IRA was depleted at the 25% rate, but the impact of loss of income from SS for those few years was much larger than the taxes saved. The Roth would be drained by about age 85. A really bad idea. Good thing I checked this.

Twists:
Remember, these four cases are really base cases.

If I continue to work after 67, I pay higher taxes but I net more, so OK. Still convert until 70.5, pay the taxes with after-tax money. Also contribute straight to the Roth. If I work until 71, our options are limited to living with RMDs or going gangbusters and converting the IRA at the top of the 25% rate.

Of course, depending on life events in any particular year, I can always take a higher or lower income, take less (until MRDs start at 70.5) or more out of the IRA and convert more or less and/or take more or less out of the Roth in distribution mode. If we can live on less than I planned here (house-sitting in Guatemala?), more can go into the Roth in conversion phase.

Being free of MRDs at some point would solve the problem of higher taxes if one of us dies before 100 (highly probable, but can happen in any of these cases). The lowest-tax case, Case B, has that trap. I have not investigated this yet, but it could make it not the lowest-tax case anymore, meaning: wipe out the IRA at all costs.

Of course, returns will not be smooth. The average ROI may not be as high as 6%. We may be forced to live on less than I planned.

The Feds could change the rules again (for example, eliminate conversion, eliminate tax-protected accounts altogether, change the marginal rates, etc.).

SS may be cut back to 78% in 2037 (we will be 90 then, if we are lucky).

By the way, if you want to try it yourself, the tax estimator at the TurboTax web site is the easier to use and gives slightly different (lower) tax estimates than the estimator I used. It is only an estimate, remember.

 

[Ed_The_Gypsy]

It may not be clear, but these are rock-bottom cases. If I work longer or have good fortune in some way, the bottom lines will improve. These are sitting-on-the-porch-watching-the-cars-go-by figures.

 

[2B]

I didn't think you were that much older than me. This is also the first time I can recall where you actually named a target retirement date.

One fly in the IRA withdrawal ointment is the much higher tax rate that will be incurred after one of the couple dies. Suddenly, any scheme to minimize or smooth out the tax rates gets derailed.

 

[Ed_The_Gypsy]

I am not intending to pack it in next year, but it was time to do some planning. The project I am on is terminating at the end of the year. There are signs that the work may continue under other auspices. I have also been told that there are other jobs I could do here, but it hasn't happened yet. The work is good, the pay is stellar, the people are easy to work with and the weather is great. (The view is great as well. I may post a photo later of the view from the window of our flat.) I hope to stick around for another year (more, if DW will allow).

Agree on the consequences of death of a partner. I have not quantified that yet, but it does suggest that the best course of action is to empty the IRA as fast as possible.

Most of these results are intuitively obvious. I did not realize how important they were until I ran the numbers, though.

And, yeah, I am older'n dirt. Early retirement passed me by a long time ago.

 

[pb4uski]

So it seems that the optimal strategy is to take as much IRA as you can at 15% since it is clear that once you start SS your SS will put you into the 25% tax bracket. That is pretty intuitive.

I don't really understand the difference between plans B and C. The words are the same as are the withdrawals (up to top of 15% bracket) but the results are different so there is some nuance of a difference that I am not getting.

While I'm much younger , from what testing I have done my optimal strategy is similar. If I live off taxable funds and do Roth conversions to the top of the 15% bracket and file/suspend SS to age 70 while DW starts spousal benefit at her FRA, I will be in the 25% bracket for about 5 years once SS starts and then drop down into the 15% bracket.

In your projections, did you increase the tax brackets for inflation? (I did in mine).

One difference that throws a twist to my plan is that I have a few years worth of capital gains in my taxable accounts that I can crystalize at 0% gain that I plan to do before starting Roth conversions.

 

[RunningBum]

B&C differ starting at 70.5. B does RMDs, while C maxes out the 25% bracket once they hit that due to SS.

Do you think the plan has the most risk if both of you live to 100 and have the higher spending, or does the loss of SS and different tax brackets offset that? I would optimize for whichever case has the most risk of your plan failing.

I don't think it matters, but in the first post was mention of "one-time" exemption for the house sale. That's no longer a one-time exemption, but there is a limit on the exemption amount. Check the new rules (not all that new, really) and see if it changes anything.

 

[Ed_The_Gypsy]

the survivor tax bomb

OK, I just checked Case B (the lowest total tax if we both live to 100) as if one of us dies at age 71 (keeping the same after-tax income). I suppose I could reduce the after-tax income by 25 or 30% but I haven't done that yet.

Total taxes to 100 rise to $117k, destroying the tax advantage of Case B. Total income to 100 drops to 78% of what it would have been if we both lived to 100. The Roth becomes exhausted at 86 years old. The IRA has only $24.6k in it at 86, so that is the only emergency fund left.

Clearly, the smart thing to do is to empty the IRA fast (Case D).

Look close and you will see the difference at age 71 (the same age I turn 70.5). In Case B, only MRDs are taken from the IRA; no withdrawals beyond that. The IRA is not drained. In Case C, at age 71, extra withdrawals are taken to the top of the 25% bracket until the IRA is drained in 3 years.

I did not increase the tax brackets for inflation. I can't predict future tax brackets. It should not change the comparison between cases as they are all on a common basis. I am not looking for absolute values, only comparisons between approaches.

I did increase after-tax income requirements for inflation and SS at a reduced rate.

 

[Ed_The_Gypsy]

Thanks, Bum. My house sale information was out of date. As you say, it doesn't matter for these calculations. It would apply to all cases and would be a bonus.

Do you think the plan has the most risk if both of you live to 100 and have the higher spending, or does the loss of SS and different tax brackets offset that? I would optimize for whichever case has the most risk of your plan failing.

Sorry, I don't understand the question. 'Spending' is after-tax income that I have increased from the first year by 3% inflation per year. At some point in all cases, that requires withdrawals from the Roth. We can economize and take less in one year. If we have an emergency, we can take out more. 'Failure' here is exhausting the Roth, which means no reserves.

 

[haha]

It becomes difficult for me to follow these things closely, as I usually would structure the whole problem differently. Let me understand-when you have succeeded in emptying your TradIRA I guess by converting some and spending some, you will then have SS as your main source of cash flow, plus an unknown amount that you may withdraw from earnings or principal in the Roth, regularly or from time to time.

Since you have always been well employed, I assume your SS plus your wife's or her spousal on your account may actually be enough for your spending needs, as you own a home in a promising, moderate cost community. In this case the Roth can just be a growing pot unimpeded by any taxation.

This ought to work fine. I previously converted up to the top of the 25%, but I started late so RMDs made that seem unwise to continue. My major concern at large conversions is that governments need money, they know exactly what and where our retirement accounts are, and if they stole 10% some year, I might prefer it to be stolen from yet to be taxed money.

Ha

 

[RunningBum]

2B had pointed out that if one of you dies, the assumptions change, and I think you've said that case D works best here. It seems that if you both make it to 100, case B works best, right? So now you have to decide whether D or B is better.

My point was, which scenario is at most risk of failing?

If there's more risk when you both make it to 100 because of higher expenses, go with B.

But if there's more risk when one of you dies early because expenses don't go down that much but the single tax rate is worse and the loss of SS income is missed, go with D.

If there's not much difference in risk, I'd probably go with the one that has less fluctuation between the two cases. D looks like the worst case for both living to 100, so if B isn't too bad if one dies earlier, go with B.

The caveat is that I didn't take too close of a look at your details, so I may be missing or misunderstanding some things.

 

[omni550]

Ed,

Did you run your various scenarios through i-ORP? Retirement Calculator - Parameter Form

omni

 

[Animorph]

I would ignore all the total taxes paid amounts. The only thing that matters is the percentage tax rate paid. Or in this case, the final portfolio value. 15% of the IRA at the start of the IRA is the same as 15% of the IRA at the end, even though the dollar amounts will be very different. You just want to minimize that tax percentage, on average, for all withdrawals.

Interesting thought on tax problem for a surviving spouse. I think if you could take additional amounts out at 15% it might be worth it. But if you make withdrawals at 25% tax rate that you could take later at 15% if both survive, then you have to balance the likelihoods of one or both spouses surviving as planned. No real winner there, just a balance.

 

[brewer12345]

All of this tells me that I should be seeking to steadily do conversions every year that I have room in the 15% bracket.

 

[kaneohe]

Ed.......if the aftertax spending is the same for all plans, isn't the figure of merit, the value of the remaining accounts (or perhaps the aftertax value of them). The amount of taxes is then just an interesting fact but not really important to you?

 

[pb4uski]

....I did not increase the tax brackets for inflation. I can't predict future tax brackets. It should not change the comparison between cases as they are all on a common basis. I am not looking for absolute values, only comparisons between approaches....

I think it will make a difference and depending on your situation, perhaps enough of a difference that it affects the decision. As you may know, by law the tax brackets will now increase for inflation rather than increase only when Congress feels like increasing them. The top of the 15% tax bracket today is $72,500 for a couple. At the same 2% rate you increase SS, it will be $80,046 when you are 71.

In my case indexing tax brackets makes a difference. If I have no indexing then beginning at age 70 I am in the 25% bracket for the rest of my life. If I index at 2%, I get back into the 15% bracket in my mid-90s. If I index at 2.5% I get back into the 15% bracket in my mid 80s and if I use 3% then I never get into the 25% bracket. YMMV.

 

[Ed_The_Gypsy]

It becomes difficult for me to follow these things closely, as I usually would structure the whole problem differently. Let me understand-when you have succeeded in emptying your TradIRA I guess by converting some and spending some, you will then have SS as your main source of cash flow, plus an unknown amount that you may withdraw from earnings or principal in the Roth, regularly or from time to time.

Since you have always been well employed, I assume your SS plus your wife's or her spousal on your account may actually be enough for your spending needs, as you own a home in a promising, moderate cost community. In this case the Roth can just be a growing pot unimpeded by any taxation.

This ought to work fine. I previously converted up to the top of the 25%, but I started late so RMDs made that seem unwise to continue. My major concern at large conversions is that governments need money, they know exactly what and where our retirement accounts are, and if they stole 10% some year, I might prefer it to be stolen from yet to be taxed money.

Ha

Spot on.

 

[Ed_The_Gypsy]

Ed,

Did you run your various scenarios through i-ORP? Retirement Calculator - Parameter Form

omni

No. I do not understand i-ORP. Too complicated for me.

 

[Ed_The_Gypsy]

I would ignore all the total taxes paid amounts. The only thing that matters is the percentage tax rate paid. Or in this case, the final portfolio value. 15% of the IRA at the start of the IRA is the same as 15% of the IRA at the end, even though the dollar amounts will be very different. You just want to minimize that tax percentage, on average, for all withdrawals.

Quite correct. There are two indexes of success/failure:
1) Is the income stream maintainable at the 3% inflation rate until the end?
2) What is the end value of the portfolio?
The total tax is only an interesting number.

Interesting thought on tax problem for a surviving spouse. I think if you could take additional amounts out at 15% it might be worth it. But if you make withdrawals at 25% tax rate that you could take later at 15% if both survive, then you have to balance the likelihoods of one or both spouses surviving as planned. No real winner there, just a balance.

Just eyeballing it, after one death, the survivor will be in the 25% bracket and nothing advantageous can be done at that point. I may look at a post-mortem IRA-emptying move.

 

[Ed_The_Gypsy]

All of this tells me that I should be seeking to steadily do conversions every year that I have room in the 15% bracket.

ABSOLUTELY! If you have enough time, move it all. I don't have enough time anymore.

This is especially important if you have a spouse. In my case, either I will convert everything to the top of the 25% bracket before 70.5 or to the top of 15% before 70.5 and to the top of 25% after 70.5 (takes 3 more years).

 

[Ed_The_Gypsy]

Ed.......if the aftertax spending is the same for all plans, isn't the figure of merit, the value of the remaining accounts (or perhaps the aftertax value of them). The amount of taxes is then just an interesting fact but not really important to you?

Yes. You've got it.

 

[Ed_The_Gypsy]

I think it will make a difference and depending on your situation, perhaps enough of a difference that it affects the decision. As you may know, by law the tax brackets will now increase for inflation rather than increase only when Congress feels like increasing them. The top of the 15% tax bracket today is $72,500 for a couple. At the same 2% rate you increase SS, it will be $80,046 when you are 71.

In my case indexing tax brackets makes a difference. If I have no indexing then beginning at age 70 I am in the 25% bracket for the rest of my life. If I index at 2%, I get back into the 15% bracket in my mid-90s. If I index at 2.5% I get back into the 15% bracket in my mid 80s and if I use 3% then I never get into the 25% bracket. YMMV.

That is good to know, P. That will help me convert at 15% before 70.5 and RMDs. I plan to do a detailed investigation with TurboTax every year after I leave this job.

 

[Ed_The_Gypsy]

All of this tells me that I should be seeking to steadily do conversions every year that I have room in the 15% bracket.

I am glad this helped someone.

 

[marc515]

In reading this thread, I understand the tax brackets are indexed for inflation and will surely increase. However, I thought the standard deduction is also indexed, effectively preventing bracket creep, and thereby keeping the effective tax rate the same?

 

[RunningBum]

In reading this thread, I understand the tax brackets are indexed for inflation and will surely increase. However, I thought the standard deduction is also indexed, effectively preventing bracket creep, and thereby keeping the effective tax rate the same?

No, what keeps the effective tax rate pretty much the same is income going up with inflation for most people, though not necessarily retirees. The standard deduction increasing actually helps you, because increasing it decreases your net taxable income.

 

[brewer12345]

ABSOLUTELY! If you have enough time, move it all. I don't have enough time anymore.

This is especially important if you have a spouse. In my case, either I will convert everything to the top of the 25% bracket before 70.5 or to the top of 15% before 70.5 and to the top of 25% after 70.5 (takes 3 more years).

We are both 40 with the bulk of our assets in traditional IRAs. I will hang it up in January and then the tax optimization begins. Since we will qualify for obamacare subsidies things will be complex, but I expect to do a lot of conversion in the next couple decades.