My Roth/HSA Reasoning

[landover]

I have to thank TromboneAl and others for starting/adding to this thread. It was on my to do list since beginning of year but I was procrastinating for long.

 

[Koolau]

I see your point, and yes, it is likely worth the paperwork. But it still leaves the TIRA with RMDs that will force a 35% tax rate, especially if I delay SS until 70.

For the moment, I have not decided for this year and I'm in the same situation at Al, which is I can control my income to near zero and do a maximum Roth conversion in the 15% bracket. But I'm vacillating on whether I want to establish yet another account for each of us.

Looking back, I wonder if it would have been better had I simply put the "matched" amount into my 401k.

At the risk (again) of sounding like the one-note song, I have to wonder if the 35% rate will stay at 35% or will it go even higher. (I just assume that lowering the 35% rate will not happen in my life time.) So, if you can convert at 15%, 25%, 28% it would seem like a comparative "bargain". I console myself with the thought that this is a "good" problem to have (though my problem isn't as big as having $2,000,000 ) It just seems that anything you can convert at less than 35% (even if it barely reduces the TIRA nest egg) is a good thing to do. As always, YMMV.

I agree with you that, given a do-over, I would not have put more into the 401(k) than it took to get the match. It was nice at the time, but, come age 70, it's going to feel like major surgery without anesthetic.

 

[RunningBum]

I see your point, and yes, it is likely worth the paperwork. But it still leaves the TIRA with RMDs that will force a 35% tax rate, especially if I delay SS until 70.

The difference between 35% and 15% on $40K is $8K. Maybe I'm missing something, but it sounds to me like you're viewing this as "all or nothing". Doing a partial conversion is far away from all, but it's sure better than doing nothing.

 

[TromboneAl]

You sir are a man of action. I have been looking at the same think and need to pull the trigger.

I just kept picturing Jan 1, 2011 and me saying "Oh shoot, I forgot to do the Roth conversion." Also, last year I didn't do a complete analysis, and I could have converted an additional $10,000 for free.

 

[haha]

After doing some calcs, it seems to me that you need to keep EVERYTHING under the 15% bracket. Otherwise, not only are you being taxed 15% on the conversion income (assuming that's the bracket your income less divs and LTCG lands), but also 15% for every dollar your divs and LTCGs push you over the 15% bracket, so effectively each dollar over the bracket costs you 30%. Is that right?

I want to be sure that I am understanding this issue correctly before 2010 expires.

RB, I cannot understand your meaning in the post quoted above. Under current law, in no case will you ever pay 30% on ltcg(s). You will either pay 0%, or 15%, depending on whether all taxable income including capital gains falls within the 15% bracket. If all taxable income is at or under the 15% bracket, your capital gains will be taxed at 0%, and your ordinary income will be taxed according to where it falls in the 10% or 15% brackets.

In no case will capital gains push your ordinary income into higher brackets. You stack blocks. Ordinary income blocks go on the bottom, and LTCG (&qual divs) get stacked above the ordinary income. As long as your ordinary income blocks do not themselves reach beyond the 15% bracket, your capital gains will not affect the tax you pay on this ordinary income. However, the amount of ordinary income can affect the rate you pay on the ltcg(s), as it can cause them to fall above the 15% bracket, below which they are taxed at 0%, into higher brackets where they are taxed at 15%.

The LTCG direct tax rate will not go above 15%- however if you have very large capital gains it can push your AGI and total taxable income to points where certain deductions, credits, etc. get scaled back or eliminated, so it can effectively raise your total taxes.

The above applies to garden variety low term capital gains, and currently also to qualified dividends. There are special categories of capital gain that do not qualfy for 0% or 15% capital gains tax.

As getting this right is important, I would welcome any comments, criticisms, or additions.

Ha

 

[RunningBum]

I want to be sure that I am understanding this issue correctly before 2010 expires.

RB, I cannot understand your meaning in the post quoted above. Under current law, in no case will you ever pay 30% on ltcg(s). You will either pay 0%, or 15%, depending on whether all taxable income including capital gains falls within the 15% bracket. If all taxable income is at or under the 15% bracket, your capital gains will be taxed at 0%, and your ordinary income will be taxed according to where it falls in the 10% or 15% brackets.

In no case will capital gains push your ordinary income into higher brackets. You stack blocks. Ordinary income blocks go on the bottom, and LTCG (&qual divs) get stacked above the ordinary income. As long as your ordinary income blocks do not themselves reach beyond the 15% bracket, your capital gains will not affect the tax you pay on this ordinary income. However, the amount of ordinary income can affect the rate you pay on the ltcg(s), as it can cause them to fall above the 15% bracket, below which they are taxed at 0%, into higher brackets where they are taxed at 15%.

The LTCG direct tax rate will not go above 15%- however if you have very large capital gains it can push your AGI and total taxable income to points where certain deductions, credits, etc. get scaled back or eliminated, so it can effectively raise your total taxes.

The above applies to garden variety low term capital gains, and currently also to qualified dividends. There are special categories of capital gain that do not qualfy for 0% or 15% capital gains tax.

As getting this right is important, I would welcome any comments, criticisms, or additions.

Ha

Yes, you stack blocks. As long as your stacks together don't go over the 15% mark ($34K for singles), the LTCG and dividends are not taxed. Ideally you push exactly to that limit. Such that your last dollar converted is taxed at 15%, and your LTCG and divs remain untaxed.

If you add to the lower bock of ordinary income by converting more of your Roth IRA, part of the upper block now pokes over that 15% line. That amount that is over the line is taxed at 15%. So if you converted $100 too much, that $100 of ordinary income is taxed at 15%, PLUS $100 of your LTCG is now taxed at 15%. That's a $30 tax on that $100 you added. That's the effective 30% rate I mentioned. If you don't have any LTCG or dividends, this doesn't apply.

I put a long post on it here: http://www.early-retirement.org/forums/f28/roth-conversion-question-53607.html#post1010935

Run it through a tax program yourself or let's see if others agree. Sue J had one addendum, that I didn't mention the personal exemption that you deduct from your income.

 

[haha]

Yes, you stack blocks. As long as your stacks together don't go over the 15% mark ($34K for singles), the LTCG and dividends are not taxed. Ideally you push exactly to that limit. Such that your last dollar converted is taxed at 15%, and your LTCG and divs remain untaxed.

If you add to the lower bock of ordinary income by converting more of your Roth IRA, part of the upper block now pokes over that 15% line. That amount that is over the line is taxed at 15%. So if you converted $100 too much, that $100 of ordinary income is taxed at 15%, PLUS $100 of your LTCG is now taxed at 15%. That's a $30 tax on that $100 you added. That's the effective 30% rate I mentioned. If you don't have any LTCG or dividends, this doesn't apply.

I put a long post on it here: http://www.early-retirement.org/forums/f28/roth-conversion-question-53607.html#post1010935

Run it through a tax program yourself or let's see if others agree. Sue J had one addendum, that I didn't mention the personal exemption that you deduct from your income.

Thanks for responding RB.

I agree with this. You have an approx $34,000 column of taxable income to fill in whatever way. If you could somehow fill the whole thing with ltcgs, you would pay no tax at all. If all of it gets filled by ordinary income (and conversions are part of this), none of your ltcg gets taxed at 0%.

For most of us who need cash flow to live, and who do not have a supply of capital losses, it still may represent a good deal. If almost all of one's ordinary income is going to be Roth conversions, and he also has considerable capital gains, he may want to consider postponing conversions.

But it still may be the cheapest long term plan to convert. It is a very complicated question with lots of moving parts.

Ha

 

[haha]

I am really glad that I asked RB to go through his reasoning, because the more I thought about it the more I realized that my plan was not optimal. Once it got through my head that I would be paying 30% on a good part of my conversion, suddenly 25% didn't look bad.

Anyway, since the taxable income from a conversion can be realized in 2011 and 2012, and since I have unusual opportunity for a large itemized deduction this year, I will take even more LTCGs, realize the CV income in 2011 and 2012, and still get a big hunk of money out of my rollover IRA and thus minimize my RMDs.

It is easy for me to either think this is too complicated and punt, or think I understand it when I don't necessarily get all the details-and some of these can turn out to be pivotal. I was up at 6:30 this morning in spite of going to bed late, because my brain was starting to wrap itself around some of the moving parts. Same thing would happen to me when I was a young boy. I would be thinking about some project and halfway through the night it might start to clear up and I would feel like I had to get it understood now!

Ha

 

[Zero]

At the risk (again) of sounding like the one-note song, I have to wonder if the 35% rate will stay at 35% or will it go even higher. (I just assume that lowering the 35% rate will not happen in my life time.) So, if you can convert at 15%, 25%, 28% it would seem like a comparative "bargain". I console myself with the thought that this is a "good" problem to have (though my problem isn't as big as having $2,000,000 ) It just seems that anything you can convert at less than 35% (even if it barely reduces the TIRA nest egg) is a good thing to do. As always, YMMV.

I agree with you that, given a do-over, I would not have put more into the 401(k) than it took to get the match. It was nice at the time, but, come age 70, it's going to feel like major surgery without anesthetic.

Here is my un-anesthetized tax surgery!

An example for a single (keeps things simpler)
Current IRA value = $1,500,000
Age to start Roth Conversion: 59.5
Age to start RMD amounts: 70
So 10 opportunities to do the Roth Conversion (RC).

For 2010, this is the headroom I have available for stacking up my blocks.
15% Bracket $34,000
Exemption $3,650
Deduction $5,700
CarryOver $3,000
Total Headroom = $46,350

These are the 2010 blocks I have to stack.
Dividends = $4,500
STCap Gain = $3,000
Roth Convert = $38.850
Total Stack Value = $46,350

Now if I take my current IRA value and assume that it will earn 5% for the next 10 years. And during that 10 years, I take out $35,000 each year as a Roth Conversion. Using Excel, here is my IRA’s performance until age 70.

2010​

$1,500,000​

2011​

$1,534,208​

2012​

$1,610,918​

2013​

$1,691,464​

2014​

$1,776,037​

2015​

$1,864,839​

2016​

$1,958,081​

2017​

$2,055,985​

2018​

$2,158,784​

2019​

$2,266,723​

2020​

$2,380,059​

Therefore, in 2020, I will be required to take an RMD of (1/24.7)= 4% of $2,380,059 = approx $100,000.
I will also have $35,000 of Social Security coming in at age 70. Gains from my taxable portfolio might also need to be taken, and so I fear a 33% tax bracket at age 70.

So what happened to the concept that putting away in a 401k would lead to paying less taxes?

The myth unraveled?

 

[Koolau]

Here is my un-anesthetized tax surgery!

An example for a single (keeps things simpler)
Current IRA value = $1,500,000
Age to start Roth Conversion: 59.5
Age to start RMD amounts: 70
So 10 opportunities to do the Roth Conversion (RC).

For 2010, this is the headroom I have available for stacking up my blocks.
15% Bracket $34,000
Exemption $3,650
Deduction $5,700
CarryOver $3,000
Total Headroom = $46,350

These are the 2010 blocks I have to stack.
Dividends = $4,500
STCap Gain = $3,000
Roth Convert = $38.850
Total Stack Value = $46,350

Now if I take my current IRA value and assume that it will earn 5% for the next 10 years. And during that 10 years, I take out $35,000 each year as a Roth Conversion. Using Excel, here is my IRA’s performance until age 70.

2010​

$1,500,000​

2011​

$1,534,208​

2012​

$1,610,918​

2013​

$1,691,464​

2014​

$1,776,037​

2015​

$1,864,839​

2016​

$1,958,081​

2017​

$2,055,985​

2018​

$2,158,784​

2019​

$2,266,723​

2020​

$2,380,059​

Therefore, in 2020, I will be required to take an RMD of (1/24.7)= 4% of $2,380,059 = approx $100,000.
I will also have $35,000 of Social Security coming in at age 70. Gains from my taxable portfolio might also need to be taken, and so I fear a 33% tax bracket at age 70.

So what happened to the concept that putting away in a 401k would lead to paying less taxes?

The myth unraveled?

Zero, thanks for the nice example. I did a similar but back of the envelope calc. on this a few years ago and that's when I started making ROTH conversions. While there are pitfalls that many have pointed out e.g., if you depend on significant LTCG - (and I don't), if you get pushed into AMT or other kinds of write-off killers, this can all get very complicated, tax wise.

Couple of points I've made before:

1) This is a "good" problem to have. I'm glad my stash is big enough to have to worry about conversions and other tax cr@p. I resent the tax codes and other BS we're forced to deal with but I often give thanks to my chosen deity that THIS is my problem and not depending on $400/mo SS to live on!!

2) If you are "lucky" even after lots of ROTH conversions, your port will grow so big, you'll STILL be paying ridiculous amounts of tax on your RMDs, This is still in the good-news-bad-news-good-problem-to-have territory.

3) I've alluded to these issues a few times before (I'm sure you're all sick of it by now) and I sure don't offer advice, but for all you Young Dreamers out there, you might consider the tenuous nature of the tax "structure" before committing to playing this game with the FEDs and IRS. As always, YMMV.

 

[Zero]

Koolau, I have only one person to blame, myself. I should have done things differently. As you clearly state, mileage may vary for others.

Interesting that I feel duped by "conventional wisdow".

 

[haha]

Interesting that I feel duped by "conventional wisdow".

If only I had been similarly duped! Oh, the pain of it!

Moe seriously, conventional wisdom is usually a bad guide to anything when it comes down to one's individual situation, as it always eventually must.

Ha

 

[RunningBum]

Zero, do you really have $1.5M or thereabouts in your 401K/IRA? Wow, what a great problem to have! Congrats!

Three observations:
1) If you convert $35K each year for 10 years, that's $350K converted at an 18% lower tax bracket, or $63K saved on taxes. It's actually a bit better since some of that conversion is taxed in the 10% bracket, and the part that offsets your deductions+exemption is free.

2) How about converting up to the top of the 28% bracket each year? Converting ~$170K each year for 10 years will bring your IRA down quite a bit. Maybe to the top of 25% is enough. Run your numbers and see how it looks. That amount over the first $350K won't be nearly as good as the 18% gain as the first part, but even if you get a 3% gain on the next million that's another $30K. I'm not sure how your numbers will actually work out but it's worth a look.

I know I talked about that hidden "30%" rate, but you're only looking at $7500 taxed at that rate before you fall back to a real 25% rate since there will be no more of the divs + ltcg block getting pushed above the $34K line.

3) Which tax bracket where you in when you contributed most of the 401K money? Remember that this was taken off the top so that's money that would've been taxed at the highest rate you were at. Now when you convert and withdraw, you are filling in the lower brackets first and finally may peak at 28% or 33%. I don't think you were duped as badly as you think you were.

However, that does make a point that if you are going to have a successful career and have a good income in retirement, it may not make that much sense to slam all you can in a 401K in your early 20s when you are in a lower tax bracket. Just getting the company match may be all you want to do. But how does one know?

 

[jdw_fire]

zero,

i took a look at your ira example with the same thought that runningbum stated i.e. that you should do some conversion at higher tax brackets. when i did this i found some things.
1) if you fill up the 25% tax bracket for those years then at in 2020 it looks like you will have less in your ira than you have now and the RMD wont push you out of the 25% tax bracket.
2) if you fill up the 28% tax bracket with conversions over that decade your ira will be almost entirely converted by 2020
3) the numbers you posted as ira balances after 2011 dont show that you are doing the converting you said you did. eg. the 2012 balance is 5% higher than the 2011 balance so no conversion was made.
and my analysis didnt increase the tax brackets for inflation which means that you would be able to convert even more than my analysis did and therefore your ira balance will be even smaller making the RMD smaller also

 

[teejayevans]

An important consideration is if you have heirs, if they are in the %15 tax bracket, and you won't be spending down your IRA, it would be just as well for them to inherit an IRA and pay 15% as oppose to you paying 25% to do the conversions.
TJ

 

[RunningBum]

An important consideration is if you have heirs, if they are in the %15 tax bracket, and you won't be spending down your IRA, it would be just as well for them to inherit an IRA and pay 15% as oppose to you paying 25% to do the conversions.
TJ

Yes, as long as you don't hit the estate tax. Then you are better off paying the income tax on the conversion if it will bring you down under that cap.

 

[Zero]

Zero, do you really have $1.5M or thereabouts in your 401K/IRA? Wow, what a great problem to have! Congrats!

Three observations:
1) If you convert $35K each year for 10 years, that's $350K converted at an 18% lower tax bracket, or $63K saved on taxes. It's actually a bit better since some of that conversion is taxed in the 10% bracket, and the part that offsets your deductions+exemption is free.

2) How about converting up to the top of the 28% bracket each year? Converting ~$170K each year for 10 years will bring your IRA down quite a bit. Maybe to the top of 25% is enough. Run your numbers and see how it looks. That amount over the first $350K won't be nearly as good as the 18% gain as the first part, but even if you get a 3% gain on the next million that's another $30K. I'm not sure how your numbers will actually work out but it's worth a look.

I know I talked about that hidden "30%" rate, but you're only looking at $7500 taxed at that rate before you fall back to a real 25% rate since there will be no more of the divs + ltcg block getting pushed above the $34K line.

3) Which tax bracket where you in when you contributed most of the 401K money? Remember that this was taken off the top so that's money that would've been taxed at the highest rate you were at. Now when you convert and withdraw, you are filling in the lower brackets first and finally may peak at 28% or 33%. I don't think you were duped as badly as you think you were.

However, that does make a point that if you are going to have a successful career and have a good income in retirement, it may not make that much sense to slam all you can in a 401K in your early 20s when you are in a lower tax bracket. Just getting the company match may be all you want to do. But how does one know?

Yes, I do have $1.5mil and it's because it's a combo of my 401k rolled into the IRA and a lump sum pension also rolled into the IRA.

I was in various tax brackets over the years and had more deductions, so I'd estimate that I was in a 20% bracket on average.

As you mention, a young person in their 20's really needs to look at this issue carefully. Especially someone in a profession where the progression is slow but sure (say engineering) where they may be able to sock away enough 401k to cause a 33%+ tax at RMD time.

 

[Zero]

zero,

i took a look at your ira example with the same thought that runningbum stated i.e. that you should do some conversion at higher tax brackets. when i did this i found some things.
1) if you fill up the 25% tax bracket for those years then at in 2020 it looks like you will have less in your ira than you have now and the RMD wont push you out of the 25% tax bracket.
2) if you fill up the 28% tax bracket with conversions over that decade your ira will be almost entirely converted by 2020
3) the numbers you posted as ira balances after 2011 dont show that you are doing the converting you said you did. eg. the 2012 balance is 5% higher than the 2011 balance so no conversion was made.
and my analysis didnt increase the tax brackets for inflation which means that you would be able to convert even more than my analysis did and therefore your ira balance will be even smaller making the RMD smaller also

Woooops, you caught me, I was doing a comparison with and without Roth Conversion and snipped out the wrong portion.

Thanks for the suggestions (you and RB), as I am clearly in need of some creative solutions.

AND, as always I feel like I am playing IRS Roulette wherein they may change the rules at any time.

Here is the IRA growth with $35,000 taken each year for Roth Conversion and a 5% growth.

2010 $1,500,000
2011 $1,538,250
2012 $1,578,413
2013 $1,620,583
2014 $1,664,862
2015 $1,711,355
2016 $1,760,173
2017 $1,811,432
2018 $1,865,253
2019 $1,921,766
2020 $1,981,104
That would reduce the RMD to $80,000.

Maybe having 5-6 children between now and 70 is the answer! OR, a market similar to the last 8 years would solve it.