Roth Conversions.

[Al in Ohio]

Just finished my Fed taxes.

I am 55 and set to retire later this year. My wife will still work, but our household income will drop by more than half. This year we were in the 25% tax bracket with an effective tax rate of 11%.

Should I consider ROTH conversions the next few years?

If so, pros and cons?


Sent from my iPad using Early Retirement Forum

[LOL!]

Sure, why not?

If you are going to pay taxes on your withdrawals at 25% anyways, you might as well convert. But it seems with you retiring later this year that this year you might be in the 15% tax bracket.

But be sure to pay the taxes on the conversion from salary or a taxable account that causes you no extra taxes to cash in. Do not pay taxes from the IRA money itself or that would be very bad (early withdrawal penalty, less money going into the Roth, etc.).

Lots of folks do multiple conversions and recharacterize ones that lose money.

As they say, run the numbers.

[pb4uski]

Another thing to know is that just because you design your conversion to bring you to the top of the 15% tax bracket does not mean that you pay 15% tax on the conversion. Ordinary income like conversions is taxed first, so the first $20k or so is covered by deductions and exemptions, the next $18k is at 10% and the remainder is at 15%. Then qualified dividends and LTCG is at 0% as long as the total taxable income is below the top of the 15% tax bracket.

The Roth conversion along with ordinary income get the benefit of the deductions, exemptions and the progressivity built into the tax brackets. I've been paying ~10% federal over the last few years.

Play with your numbers in Taxcaster.

[bingybear]

you really should compare what you'd estimate the marginal tax rate will be at RMD time and what the marginal tax rate will be for a conversion that you do now. Also remember that you will likely have SS at RMD time. If you can pay lower taxes now and reduce your RMD tax rate overall, then do it. The above comments are should be followed about using after tax $ to pay the taxes.
Much of the decision depends on your situation.

[Whisper66]

May want to make some runs with the free online tool, Optimal Retirement Planner - Parameter Form . The tool tries to optimize how much money you can use over your lifetime. One of the check boxes allows Roth conversions to be considered and the tool will recommend them if they help optimize your overall lifetime spend. I think the free online tool America's Top Financial Planning Tool - And It's Free! | ESPlannerBasic is supposed to provide similar information but I wasn't able to figure it out. Both tools are used by many on this board.

I used I-ORP to study if conversions were worthwhile. In my situation, doing conversions for several years is optimum due to lowering my future taxes ...especially when I am collecting SS and having to also pay taxes on Required Minimum Distributions. I've run all my calculations using current income tax assumptions....but I assume if they change, they will only get higher which makes conversions more attractive in my situation.

Lots of good discussions on this forum on if Roth conversions make sense and how much to convert if one chooses to do it. Opinions vary quite a bit.

[sengsational]

I think the reasons opinions vary is because individual situations vary.

I second the idea for doing analysis on i-orp. Get everything in there as tidy as you can, then run once with the Roth checkbox on, and once with it off. Pay attention to the graph near the bottom of results that shows how much of your income is taxed at the various percentages. Many people are surprised that it recommends edging out of low tax brackets in earlier years, but it does that to prevent high taxes later, when you would be taking RMD's from your IRA's.

If you read the linked paper on the i-orp form, you will learn that in many cases (many situations), doing Roth conversions isn't really a very big deal in the big picture. In other words, you could do something else (like retire in a no income tax state) would make a huge difference when compared to fiddling around with Roth conversions. Even though Roth conversions don't make a very big difference in my situation, I'm still doing them since it's a simple thing to do.

[samclem]

For most of us, the modeling of the results of Roth conversions is highly sensitive to the assumed portfolio return: If we assume the markets will produce great returns, then we'll be in high tax brackets later on and doing Roth conversions earlier would reduce taxes. But if we assume instead that the markets produce low returns, then the Roth conversions may produce no net benefit AND we'll have paid a bunch of money in taxes earlier that is no longer in our portfolio. If investment results are really that poor, we'll probably have some real need for those dollars, but they will be gone--paid to the IRS.

In a nutshell: If my investments do well over the coming decades, I'll be in good shape and any extra taxes won't be a big concern. If my investments do poorly, then I'll miss that extra money given to the IRS. So, in consideration of this, I will be conservative with these Roth conversions in the years prior to my RMDs. I'll do some, but I sure won't be doing any conversions above the 15% bracket (where we are now, and will probably be for a LONG time).
One other consideration for couples--look at the tax tables for a single person, and at the reduced standard deduction, etc. A survivor can easily find him/herself paying taxes at very high rates. This is an argument for doing >more< Roth conversions, especially if the tax bite would significantly impact the survivor's quality of life.

It's important to keep the ultimate goal in mind. For most if us, the ultimate goal is not "maximizing the chance of having the most money when we die." Instead the goal is "maximize the chances that our spending power lasts as long as we do, and minimize the chances for a disastrous outcome." They can require very different actions.

[Al in Ohio]

Wow. Great suggestions so far. Thank you all. I need to download some programs this weekend!


Sent from my iPhone using Early Retirement Forum

[pb4uski]

Quote:

Originally Posted by samclem For most of us, the modeling of the results of Roth conversions is highly sensitive to the assumed portfolio return ...

I think you can get a reasonable idea of the impact using Excel's FV function (or doing a simple model in Excel). For example, if you are 60 and had $1 million in tax-deferred and can do $75k a year of roth conversions you can calculate your tax-deferred balance at age 70 as =fv(rate,10,70000,-1000000) and do a sensitivity table of the result at different rates of return.

At 5% your age 70 tax-deferred balance with $75k annual Roth conversions would be about 42% of what it would be without Roth conversions and your RMDs would be commensurately lower as well. At 7% and 9% the percentages are 47% and 52% respectively so the result isn't terribly sensitive to changes in assumed rates of return.

If I assume that the age 60 to 70 Roth conversions are taxed at 10% (my recent experience) and RMDs are taxed at 25% then that is $160k - $200k in (nominal) tax savings at 5% and 9%, respectively. Not chump change to me.

[hnzw_rui]

Quote:

Originally Posted by samclem It's important to keep the ultimate goal in mind. For most if us, the ultimate goal is not "maximizing the chance of having the most money when we die." Instead the goal is "maximize the chances that our spending power lasts as long as we do, and minimize the chances for a disastrous outcome." They can require very different actions.

+1. Another consideration is the possibility or large medical and/or long-term care bills later in life. That's potentially 0% tax going in and 0% tax coming out. Should greatly help if you ever find yourself in a nursing home that costs $100K/year.

[Senator]

Regardless of your tax rate now, odds are, tax rates of the future will be higher.

Convert a bit each year.

[pb4uski]

Quote:

Originally Posted by samclem .....It's important to keep the ultimate goal in mind. For most if us, the ultimate goal is not "maximizing the chance of having the most money when we die." Instead the goal is "maximize the chances that our spending power lasts as long as we do, and minimize the chances for a disastrous outcome." They can require very different actions.

I see them as similar in that the risk is living long and if you make choices to have the most money when you die don't you similarly minimize the likelihood of running out of money if you have adverse experience? I'm not certain that they are one and the same but they are certainly similar.

[samclem]

Quote:

Originally Posted by pb4uski I think you can get a reasonable idea of the impact using Excel's FV function (or doing a simple model in Excel). For example, if you are 60 and had $1 million in tax-deferred and can do $75k a year of roth conversions you can calculate your tax-deferred balance at age 70 as =fv(rate,10,70000,-1000000) and do a sensitivity table of the result at different rates of return. At 5% your age 70 tax-deferred balance with $75k annual Roth conversions would be about 42% of what it would be without Roth conversions and your RMDs would be commensurately lower as well. At 7% and 9% the percentages are 47% and 52% respectively so the result isn't terribly sensitive to changes in assumed rates of return.

I appreciate the clean simplicity of the approach, but:
-- Your starting point (5%) is already on the high end of real returns I'd expect for the next 10 years. So, we'd need to start lower. And:
-- Any comparison would need to include the value (incl growth) of the funds >not< paid to the IRS if we don't do the conversion. So the value of the tax-free portion of the portfolio only tells part of the story--the pre-tax portion counts, too.


Off topic: And an observation about RMDs: Yes, they are "forced" and can drive up taxable income in retirement. But, compared to what? Aren't we planning to spend the money eventually? Is the RMD withdrawal rate considerably higher than the rate I would have preferred? (For me--no).

If the tax rate paid today equals the tax rate I would pay in the future, then the Roth conversion is a wash--it produces no improvement in my spendable account balance. But it the taxes paid early >do< decrease my flexibility and the ability of my portfolio to survive a bad market spell.
Obviously, if the tax rate in the future is higher then I would have been better off to pay earlier at the lower rate. But if the higher rate is due to having an unexpectedly big portfolio throwing off big RMDs--then I'm rolling in dough and the taxes are a minor consideration.

[pb4uski]

At lower rates of return the benefits are higher because more of the account proceeds are being taxed at a lower rate. For example, at 3% the account is 36% and the savings are still $140k.

What I suggested is easier to understand than a complicated model (which I have done for myself) but give one a sense of the direction, magnitude and sensitivities of doing Roth conversions vs not.

While I concede that you need to consider the value of growth of the taxes paid you also need to consider the value of not ever again paying taxes on the amounts converted. Both are second order effects that we know offset directionally and trying to include them just muddies the waters but I concede that both should be included in a comprehensive model but that is not the objective of the simple calculation.

Whether and when the money in the tax-deferred account or in the Roth account is spent isn't particularly relevant to the decision.

I agree that it the tax rate today is the same as the future then the whole thing is much ado about nothing. That might be true of some people with significant pension income of really significant wealth such that we are just talking about being in the 25% and 28% bracket... a 3% spread vs the difference between 15% and 25%. I think Roth conversions work best where the difference in taxes is quite significant. You can get a fair idea about your future tax bracket by looking at your taxes today if your pensions and SS were on line and you had to take RMDs since SS, RMDs and tax brackets grow with inflation.

While I agree that if investment experience is good and RMDs pushing you into a higher tax bracket that you have the resources to pay the taxes, but I would prefer to do conversions and have those savings go to my kids than to the government.

[samclem]

Quote:

Originally Posted by pb4uski I see them as similar in that the risk is living long and if you make choices to have the most money when you die don't you similarly minimize the likelihood of running out of money if you have adverse experience? I'm not certain that they are one and the same but they are certainly similar.

Sometimes they are the same, sometimes they aren't. A thought experiment: You are offered two options for your retirement savings:
1) Bet everything on one roll of a fair die. Pick any number, and you'll get paid 10:1 odds if you are right, lose your money if wrong.
2) Don't bet

Clearly, from a net expected value POV, the first option is best. Your chance of picking the right number is 1:6, but you are being paid 10:1, it's a no-brainer. This is clearly the best way to maximize the expected amount of money, even with the possibility of losing all of it.

Most of us would choose option 2. It doesn't produce the highest expected outcome, but still we would choose it. That's because our objective isn't to maximize our expected outcome, but to maximize chances of making our spending power last through our living years. Not betting (and continuing to invest our money as normal) does that better than Option 1.

There are variations of this, but the marginal utility of money decreases as we have more of it. So, for most f us, if our total retirement income was a rock-solid $3K per month (inflation-adjusted) forever, we wouldn't take a 50:50 bet where we could lose $1500 per month even if the gain would be $2000 per month. The "negatives" of trying to get by on $1500/mo are much higher than the "positives" of a new monthly spending of $5000/mo.

The decision to delay SS is an obvious place where this dynamic comes into play, but this Roth conversion question is another one. Trying to reduce taxes if our portfolio "hits one out of the park" may not make sense if it comes at the risk of reducing available funds if our portfolio later does poorly and is barely keeping the heat on in the winter.

[samclem]

Quote:

Originally Posted by pb4uski Whether and when the money in the tax-deferred account or in the Roth account is spent isn't particularly relevant to the decision.

I think the main point of the exercise is to maximize the likely utility of the money (not the absolute number of dollars). $1 in a small account has higher utility than $1 in a much larger account, so we get more "bang for the buck" by reducing the impact of a poor string of investment returns even if it costs more than a dollar if the future returns are great.

Quote:

Originally Posted by pb4uski . . .but I would prefer to do conversions and have those savings go to my kids than to the government.

Yes, but that is subject to the same considerations. If it's a traditional IRA and the RMDs eventually put your kids in a higher bracket, then maybe in retrospect it would have been better for you to pay early. But if not, they'd probably appreciate having the extra dough in the account, you gained nothing for them by paying early when you converted. And if we are talking about "regular" after-tax money (people sometimes urge people to use these funds to pay taxes on Roth conversions), then your kids would clearly have been better off if the conversion hadn't been done (because they get a stepped-up basis on this after-tax money)

[pb4uski]

No betting going on here... it is just a simple tax rate arbitrage game.

[gauss]

If all other things are equal, if you are filing Married Filing Jointly, I say Roth convert sooner rather than later to take advantage of the Married tax brackets, standard deduction etc, which are cut in half for single filers.


None of us knows the day of our demise.....


-gauss

[pb4uski]

I was a bit curious on this so I set up a scenario of a 60 yo retired couple with a $1 million tIRA and $100k in taxable funds. One scenario is to do nothing. The other is to do $75k in Roth conversions annually and pay 10% federal tax (about what I am converting and paying and the 10% is a blend of deductions, exemptions, the 10% and 15% tax brackets). At age 70, the account values are converted to an after-tax amount at 25% as the retired couple has started SS, pensions, etc and is subject to RMDs.

The analysis suggests that at a 5% rate of return that the couple and their heirs come out about 10% ahead by doing Roth conversions. The advantage increases at lower rates of return because a higher percentage of funds is converted at lower rates and the inverse at higher rates of return. See sensitivity table.

Obviously YMMV.

		tIRA	After-tax	Roth	Total		tIRA	After-tax	Total
Beginning value @ age 60		1,000,000	100,000	-	1,100,000		1,000,000	100,000	1,100,000
Rate of return	5%
Roth conversions	10%	(75,000)	(7,500)	75,000	(7,500)				-
Term in years	10
Nominal Future Value @ age 70		685,553	68,555	943,342	1,697,450		1,628,895	162,889	1,791,784
Future taxes	25%	171,388					407,224
After-tax value		514,165	68,555	943,342	1,526,062		1,221,671	162,889	1,384,560
Extra to heirs - $					141,501
Extra to heirs - %					10.2%

Sensitivity table
0%	13.2%
2%	11.9%
3%	11.3%
4%	10.7%
5%	10.2%
6%	9.7%
8%	8.9%
10%	8.1%

[hnzw_rui]

Quote:

Originally Posted by pb4uski I was a bit curious on this so I set up a scenario of a 60 yo retired couple with a $1 million tIRA and $100k in taxable funds. One scenario is to do nothing. The other is to do $75k in Roth conversions annually and pay 10% federal tax (about what I am converting and paying and the 10% is a blend of deductions, exemptions, the 10% and 15% tax brackets). At age 70, the account values are converted to an after-tax amount at 25% as the retired couple has started SS, pensions, etc and is subject to RMDs. The analysis suggests that at a 5% rate of return that the couple and their heirs come out about 10% ahead by doing Roth conversions. The advantage increases at lower rates of return because a higher percentage of funds is converted at lower rates and the inverse at higher rates of return. See sensitivity table. Obviously YMMV.

Question, for your model, was the couple drawing from the tax deferred portfolio at the same time from age 60-70 for living expenses? Because that alone could make super size Roth conversions unnecessary. Indeed, very large Roth conversions could be dangerous if someone is relying on their portfolio for majority of living expenses.

Definitely a YMMV scenario, I would think.