Retirement Tax Planning - Income Optimization?

[PatrickA5]

You are right that there is no 8.5% tax bracket, but you are wrong in using ONLY the marginal tax bracket.

8.5% is the effective rate that I paid on the $386k of Roth conversions that I did over the last 7 years... caalculated as (tax after conversion less tax before conversion) divided by the conversion amount.... and is appropriate since the conversion crosses a number of tax brackets.... some is 0% because it is offset by the standard deduction, some is at 10% and the rest is at 12% (was 15%).

I agree with this. You really have to look at the effective tax rate for your own situation and compare effective tax rates now (during conversions) and later when RMD's kick in.

In my own situation, I'm in the marginal 12% bracket now and when I'm taking RMD's I'll be in the marginal 22% bracket. BUT, very little of my income (at age 70) will be in that 22% bracket, the vast majority will be taxed at an effective rate of less than 10%. You really have to look at the actual taxes paid, not just the marginal rate of the last taxed dollar.

I should note, that I won't have pensions or other income other than retirement accounts and SS. I hope to have a decent mix of SS, Taxable, Tax Deferred, and Tax Free, that I can more or less manipulate my withdrawals to pay the least tax possible each year. At least that's the plan.

 

[corn18]

You are brilliant. I have found an issue with my spreadsheet that allows my Roth balance to go negative, which should not happen. That would require withdrawals from my 401k to keep that from happening (and a taxable event that I am not accounting for). That might explain why the taxes are less. Will explore a fix and report back.

Found an error where I allow my Roth to go negative instead of drawing additional (and taxable) amounts from my 401k. That left out some taxes.

Here are the new numbers to age 80:

$250,780 100% conversion
$288,929 0% conversion

The difference is the tax on the 401k withdrawals needed to cover expenses before RMD's. I wasn't accounting for the tax on that amount. Oops. I think I have too many spreadsheets and variable. $1.8M with a paid off house = FAAAT FIRE, that's the only number I need.

 

[PatrickA5]

But don't forget, someone still needs to pay the tax on the money left in the IRA in the 0% conversion case, unless you bequeath it to charities. Some may say their heirs should be grateful for anything they get, taxed or not. I think it's foolish not to do some estate planning to reduce heirs taxes. Also, you died with money left in a tax deferred account that you couldn't spend from, so it's not like you optimized your own situation.

If your heirs seem likely to be in a lower tax bracket than you are in retirement, it may make sense to let them pay the taxes. But now they have to withdraw it all in 10 years, so that will likely bump up their tax bracket if the amount left is significant.

Personally, I've done a lot of estate planning, but I might have to draw the line on trying to figure out the tax implications of my bequest to each of my 3 kids (and grandkids). I'm having a hard enough time trying to figure out how much conversions help me (and DW)

 

[RunningBum]

Personally, I've done a lot of estate planning, but I might have to draw the line on trying to figure out the tax implications of my bequest to each of my 3 kids (and grandkids). I'm having a hard enough time trying to figure out how much conversions help me (and DW)

My thought on that is to just use my own tax rate for the duration of the deferred tax account. That way if I live a long, long time, I've optimized my situation. And if I don't, I've at least acknowledged that they will have some tax on it and converted appropriately. It may or may not be accurate, but at least I haven't totally ignored the inherited tIRA tax implications, without making my calculations any more difficult than it already is.

 

[RunningBum]

All of this "optimization" discussion is good, but people need to remember: Don't let the tax tail wag the investment dog, and that generally when you have to pay more taxes it means you have been successful. Don't lose site of that while one optimizes their tax situation.

When it comes to Roth conversions, it's really strictly a tax play, isn't it? It's just shuffling money from one account to another to try to reduce taxes. In this case the tax tail is the only factor in play.

 

[pb4uski]

+1 When I do my Roth conversion it goes into the same ticker that it was in in my tIRA... so no real investment implications.

And you are right.... it is principally a tax arbitrage play, with a minor second-order-effect tax benefit if you use taxable funds to pay for the tax in that that money is tax free from that point onward. I did an example recently and that minor benefit at the end of 30 years at a 7% investment earnings rate and 20% tax rate was about 7%.

 

[SevenUp]

My spreadsheet calculates taxes by year out to age 100 and accounts for % SS taxed (it is comprehensive). This tax question is easily answered for my particular situation:

Convert up to top of the 12% and then 15% tax bracket until age 70 (in 2019 dollars):

$444,719 total taxes paid

Do no conversions and let RMDs kick in @ 70:

$626,514 total taxes paid

That is not chump change and makes my decision easy.

You might want to look at "amount spendable after tax" rather than "taxes paid."

E.g., Roth will usually look better than traditional if one uses "absolute amount of tax paid" as the measure, because a smaller amount of tax is paid at contribution because the contribution amount is relatively small. When, years later, a traditional withdrawal is made, the absolute tax amount may be higher (due to investment growth and corresponding larger withdrawal amount) but what matters is the amount left after tax.

 

[SevenUp]

In my own situation, I'm in the marginal 12% bracket now and when I'm taking RMD's I'll be in the marginal 22% bracket. BUT, very little of my income (at age 70) will be in that 22% bracket, the vast majority will be taxed at an effective rate of less than 10%. You really have to look at the actual taxes paid, not just the marginal rate of the last taxed dollar.

Correct, one should not look at the marginal rate on a single dollar unless one is looking at a decision involving only that single dollar.

Given the situation you describe, however, it is indeed the 22% marginal in the future that you should compare with the 12% today.

That's because even if you do nothing now you will be paying 22% on some portion of your income. If you contribute to traditional now and save 12%, your future traditional balance and thus your RMD will increase (compared with doing nothing now), causing you to pay 22% on that increase.

Does that make sense?

 

[corn18]

When it comes to Roth conversions, it's really strictly a tax play, isn't it? It's just shuffling money from one account to another to try to reduce taxes. In this case the tax tail is the only factor in play.

It really is that simple.

 

[PatrickA5]

Correct, one should not look at the marginal rate on a single dollar unless one is looking at a decision involving only that single dollar.

Given the situation you describe, however, it is indeed the 22% marginal in the future that you should compare with the 12% today.

That's because even if you do nothing now you will be paying 22% on some portion of your income. If you contribute to traditional now and save 12%, your future traditional balance and thus your RMD will increase (compared with doing nothing now), causing you to pay 22% on that increase.

Does that make sense?

That makes sense, but I'm not trying to decide if I want to contribute now.

I'm trying to decide if I want to make conversions into the 22% bracket now when potentially only a small amount of RMD's will be taxed at 22% in the future. I suppose if I had other income (non RMD) that takes up the standard deduction and the 10% and 12% brackets when I'm 70, then it would be a no doubter to convert money in the 22% bracket now. That's not the case. We'll only have SS. The standard deduction and the 10% bracket will take care of most of the Social Security income. A large chunk of our RMD's will be taxed at 12% - even if we do nothing.

Of course the widow/widower tax rate is a different problem to figure out.

 

[Winemaker]

Using the Retiree Portfolio Model v.18, our portfolio ends up $587,000 better if no conversion, and that's converting $50,000/yr for the next 9 years. I'm 61, plan calls for our demise at 95. It would cut my RMDs $685,000 and save us $141,000 Federal tax over the next 34 years.

If I covert $100,000/yr for the next 9 years, the unconverted portfolio is $1,209,000 better. Our RMDs would be cut $1,371,000 and $271,000 taxes on that, over 34 years.

The no conversion portfolio at 95 would still be close to 8 figures either way. So we either start blowing the dough or the kids/grandkids will have get to receive it over 10 years

 

[SevenUp]

I'm trying to decide if I want to make conversions into the 22% bracket now when potentially only a small amount of RMD's will be taxed at 22% in the future. I suppose if I had other income (non RMD) that takes up the standard deduction and the 10% and 12% brackets when I'm 70, then it would be a no doubter to convert money in the 22% bracket now. That's not the case. We'll only have SS. The standard deduction and the 10% bracket will take care of most of the Social Security income. A large chunk of our RMD's will be taxed at 12% - even if we do nothing.

Probably all agree that conversions to the top of the 12% bracket are worth doing. Now the question becomes "will the conversions to the top of the 12% bracket be enough to reduce the RMD bracket below 22%?"

If "yes", then stop. One could quibble over that portion of the conversion at 12% that is reducing RMDs also at 12%, but assuming "the tie goes to Roth" due to possible future bracket rate increases is not unreasonable.

If "no", then one could ask "how far into the 22% bracket do I need to go with a conversion before the RMD bracket would drop below 22%?" and proceed accordingly. Again, assuming the tie goes to the Roth is not unreasonable.

And the use of "bracket" may be an oversimplification, e.g., due to the way Taxation of Social Security benefits, IRMAA, etc., works. It's really the marginal tax rate on the amounts in question that matter. The tax rates on the portion of RMDs unaffected by conversion choices are irrelevant.

 

[retiredunder50]

I stumbled across these 2 recently posted articles with heavy analysis that may (or may not) provide some food for thought in this thread.



https://earlyretirementnow.com/2019...curity-tax-torpedo-roth-conversion-tightrope/


https://earlyretirementnow.com/2019...tirement-without-paying-federal-income-taxes/

 

[PatrickA5]

Probably all agree that conversions to the top of the 12% bracket are worth doing. Now the question becomes "will the conversions to the top of the 12% bracket be enough to reduce the RMD bracket below 22%?"

If "yes", then stop. One could quibble over that portion of the conversion at 12% that is reducing RMDs also at 12%, but assuming "the tie goes to Roth" due to possible future bracket rate increases is not unreasonable.

If "no", then one could ask "how far into the 22% bracket do I need to go with a conversion before the RMD bracket would drop below 22%?" and proceed accordingly. Again, assuming the tie goes to the Roth is not unreasonable.
.

With DW still working, we don't have much space in the 12% bracket to convert. She'll retire in a couple of years which might give us a lot of 12% bracket to play with at that time. We will probably have 5 or 6 years to do maximum 12% conversions after she retires. This might be enough to get us out of the 22% bracket at RMD time. If it is, then I wouldn't want to do 22% conversions now.

And, there is a possible inheritance that could throw a money wrench into the calculation. Also, when we plan on taking SS. Right now the calculators show me taking at 62 (next year) and her at 70, but obviously it might make more sense for me to wait and concentrate on conversions.

I'm probably spending too much time thinking about this stuff. I imagine the overall net worth change isn't enough to get too worked up over.

 

[Midpack]

My spreadsheet calculates taxes by year out to age 100 and accounts for % SS taxed (it is comprehensive). This tax question is easily answered for my particular situation:

Convert up to top of the 12% and then 15% tax bracket until age 70 (in 2019 dollars):

$444,719 total taxes paid

Do no conversions and let RMDs kick in @ 70:

$626,514 total taxes paid

That is not chump change and makes my decision easy.

How much does it alter your ending balance?

 

[tenant13]

That just baffles me. Do you ignore price tags and money issues in everything you do? This is such a simple thing to do to save yourself 100K or more (I'm sure this varies from case to case) I don't know why you would ignore it. Why would you even spend time on a tax planning & income optimization thread if you just don't care? And how can you be sure you have enough? What if you need extended time with in home nursing care, for example?

Of course I don't ignore the prices but to answer your questions decisively I would have to know the date of my death. Without it, it's just a guessing game. It can be anywhere from 5-6 years from now (my father and his side of family is plagued by cancer) or 30 years from now preceded by a severe case of dementia (my mother, and her mother and... well, you get the picture). So while it is possible I could save some money if I live long enough, the likelihood of that is debatable. Instead I choose to look at my budgets and project them into the future. If the financial software tells me I can hit my numbers until 90 I take it from there and make converting decisions based on my current spending needs. So if I can spend a little bit more now (when I'm 56 and healthy) by not paying conversion taxes up to 22-24% bracket I will probably do just that.

As to LTC? Having literally nobody who I could trust with making health and financial decisions for me, I may just skip it altogether...

 

[RunningBum]

Of course I don't ignore the prices but to answer your questions decisively I would have to know the date of my death. Without it, it's just a guessing game. It can be anywhere from 5-6 years from now (my father and his side of family is plagued by cancer) or 30 years from now preceded by a severe case of dementia (my mother, and her mother and... well, you get the picture). So while it is possible I could save some money if I live long enough, the likelihood of that is debatable. Instead I choose to look at my budgets and project them into the future. If the financial software tells me I can hit my numbers until 90 I take it from there and make converting decisions based on my current spending needs. So if I can spend a little bit more now (when I'm 56 and healthy) by not paying conversion taxes up to 22-24% bracket I will probably do just that.

As to LTC? Having literally nobody who I could trust with making health and financial decisions for me, I may just skip it altogether...

No, you really don't need to know the date of your death. All of these decisions have to be made with predictions, assumptions, and best guesses on your life span, investment returns, future tax rates, future spending needs, etc. But for conversions, death is less of a factor since the deferred tax liability lives on.

With all the uncertainties, I actually take comfort that I'm removing the uncertainty of future tax rates on my tIRA as I convert it to a Roth. So I'm simplifying my financial planning.

There's no reason you can't convert now and still spend more now. I've said this before but you keep ignoring it. Why is that? Conversions work best if you can pay the taxes from outside (taxable account), but even if you pay the tax with the conversion money you can still come out ahead. Conversions and spending aren't and shouldn't be related. Taxes from conversion should not be part of your yearly budget. Treat it as a special expense.

 

[Midpack]

^^ There's no right answer...

There are dozens of significant unknowns*, and the right answer for each of us depends almost entirely on our assumptions and expectations. And you can be sure that some of your assumptions will prove to be wrong no matter who you are - the .01% that get them all right will be strictly by chance.

*Longevity, sequence of returns, sequence of inflation, current & future tax brackets, tax rate changes, Soc Sec changes, RMD changes, Roth changes, health care costs, estate tax laws, unexpected expenses, and many more.

 

[googily]

As a still-working widow whose tax-deferred monies make up 81% of her holdings, I would tell people that a little bit of converting will go a long way for whichever one of you survives the other. You may not both die within a day of each other at 90.

In the other thread I declared I was not going to bother with conversions while still working. The problem is, it is to my benefit to get some of this out of tax-deferred one way or the other. Since I can withdraw from the IRA I inherited from DH without penalty even before I hit 59.5, why not just withdraw some as long as I stay within my current (24%) bracket? Like it's a salary increase! Of course! But then my next thought is, well, I have money in my after-tax brokerage account, so I don't really NEED the extra money, and I have to pay taxes one way or the other on that "salary", so why not just convert it? Then I will have money to help manage my income better when I'm receiving two pensions, survivor SS at 60, my SS at 70....

And given how much the levels of the tax brackets might change after 2025 (going from $161k for the top of 24% to $94k for the top of 25% if they purely revert), the amount of space to stay in 25% shrinks considerably.

It's a problem I'd much rather have than being a widow trying to scrape through on her own, but I can never really settle on the right answer. Which is why I keep changing my tune.

 

[corn18]

^^ There's no right answer...

There are dozens of significant unknowns*, and the right answer for each of us depends almost entirely on our assumptions and expectations. And you can be sure that some of your assumptions will prove to be wrong no matter who you are - the .01% that get them all right will be strictly by chance.

*Longevity, sequence of returns, sequence of inflation, current & future tax brackets, tax rate changes, Soc Sec changes, RMD changes, Roth changes, health care costs, estate tax laws, unexpected expenses, and many more.

And the cool thing is, there is no wrong answer either, as long as you don't take it to the extreme. If I converted my entire 401k in year one, that would be a wrong answer. But if I convert none of it until RMD's, it won't kill us. And I can convert some of it before RMDs and it won't kill us. We won't know the "right" answer until we are both dead. Then I'll let my heirs decide if we were right.

My widowed MIL is taking RMDs that could have been avoided had her husband done some conversions prior to RMDs. I guess he chose not to or didn't know he could. She hates the RMD and resulting taxes, but her portfolio is big enough to not matter. She is doing fine. So he was not right or wrong in not doing conversions.

 

[PatrickA5]

And the cool thing is, there is no wrong answer either, as long as you don't take it to the extreme. If I converted my entire 401k in year one, that would be a wrong answer. But if I convert none of it until RMD's, it won't kill us. And I can convert some of it before RMDs and it won't kill us. We won't know the "right" answer until we are both dead. Then I'll let my heirs decide if we were right.
.

+1

 

[PatrickA5]

I think DW and I finally came to a decision on conversions for 2019. We aren't doing any! Instead we're taking withdrawals up to the AOP cutoff. Our cash stash is the lowest it's been in years and needs some replenishing. Our retirement accounts went up a lot this year, so we'll take a small percentage off the table. Withdrawals aren't quite as tax efficient as conversions, but they basically accomplish the same purpose. We'll probably start back up with the conversions next year.

 

[pb4uski]

+1 In both cases the tax-deferred balances are reduced and the tax torpedo is reduced a bit.

That is what I did in 2018 and some in 2019... but I'm starting to think that a better strategy is to do 100% Roth conversions but then just spend from the Roth... a hair more tax efficient since interest/growth on the funds is tax-free.

So in your case do the Roth conversion which takes some off the table as you want (assumes it is MM in the Roth) and the cash in the Roth is just as good, and arguably a hair better than cash in a taxable account since the interest is tax-free.

Once I got over my aversion to "never touch" the Roth it became a lot easier to think about.

 

[PatrickA5]

+1 In both cases the tax-deferred balances are reduced and the tax torpedo is reduced a bit.

That is what I did in 2018 and some in 2019... but I'm starting to think that a better strategy is to do 100% Roth conversions but then just spend from the Roth... a hair more tax efficient since interest/growth on the funds is tax-free.

So in your case do the Roth conversion which takes some off the table as you want (assumes it is MM in the Roth) and the cash in the Roth is just as good, and arguably a hair better than cash in a taxable account since the interest is tax-free.

Once I got over my aversion to "never touch" the Roth it became a lot easier to think about.

Good idea. Like you said if I can get over the "never touch" the Roth. Hard to do!

 

[corn18]

Good idea. Like you said if I can get over the "never touch" the Roth. Hard to do!

This is a big hangup for me as well. This chart bugs me because everything is going into my Roth and my taxable is going to zero too fast:

But I am getting over it. Having 100% of our retirement savings in a Roth is not a bad thing. Roth cash spends the same as taxable cash. At least that's what I keep telling myself.