Vanguard Paper on Roth Conversion Analysis

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Svensk Anga
Posts: 437
Joined: Sun Dec 23, 2012 5:16 pm

Vanguard Paper on Roth Conversion Analysis

Post by Svensk Anga » Mon Jul 23, 2018 9:41 pm

Vanguard just released a paper suggesting a better metric for evaluating Roth conversions, relative to the comparison of current year to future year marginal tax rates. They call it BETR, Break-Even Tax Rate. Paper is here:

https://advisors.vanguard.com/iwe/pdf/I ... IL:ET:2017

This metric accounts for the fact that using funds from a taxable account to pay the conversion taxes avoids some tax drag over the years (preferably decades) between conversion and Roth withdrawal. It makes conversions somewhat more attractive. They also point out that it can help justify conversion of IRA's that have some basis. Also, it might help one decide to fully convert tIRA's to enable back-door Roths in future years.

Those on the fence about conversions should have a look.

Miriam2
Posts: 2281
Joined: Fri Nov 14, 2014 11:51 am

Re: Vanguard Paper on Roth Conversion Analysis

Post by Miriam2 » Mon Jul 23, 2018 11:57 pm

Svensk Anga wrote: Vanguard just released a paper suggesting a better metric for evaluating Roth conversions, relative to the comparison of current year to future year marginal tax rates. They call it BETR, Break-Even Tax Rate. Paper is here:

https://advisors.vanguard.com/iwe/pdf/I ... IL:ET:2017
Thank you for this info!
Is there a link or window where we can find the Vanguard research or white papers?

User avatar
Eagle33
Posts: 80
Joined: Wed Aug 30, 2017 3:20 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by Eagle33 » Tue Jul 24, 2018 12:08 am

Miriam2 wrote:
Mon Jul 23, 2018 11:57 pm

Is there a link or window where we can find the Vanguard research or white papers?
link to Vanguard papers
https://investor.vanguard.com/investing ... t-research
Rocket science is not “rocket science” to a rocket scientist, just as personal finance is not “rocket science” to a Boglehead.

User avatar
Cyclesafe
Posts: 658
Joined: Wed Dec 31, 2014 1:03 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by Cyclesafe » Tue Jul 24, 2018 9:38 am

I get how paying taxes on a conversion from a taxable account preserves tax-free space, allowing greater tax-free compounding, but to claim extra benefit when funds come from a tax-inefficient taxable account really takes credit for solving a problem which should have been addressed separately - i.e. improving the tax efficiency of the taxable account in the first place.

For an investor for whom muni's make sense and for whom equity in taxable are in index funds that shed minimal dividends and capital gains, this analysis makes little difference. The overwhelming factor in whether a conversion makes sense remains the investor's assumption that his/her marginal tax rate will be higher in the future. If the investor wants to act on this belief and convert, then paying taxes from non-tax deferred sources continues to be the way to go - cash flow needs permitting.

cherijoh
Posts: 4945
Joined: Tue Feb 20, 2007 4:49 pm
Location: Charlotte NC

Re: Vanguard Paper on Roth Conversion Analysis

Post by cherijoh » Tue Jul 24, 2018 9:50 am

Cyclesafe wrote:
Tue Jul 24, 2018 9:38 am
I get how paying taxes on a conversion from a taxable account preserves tax-free space, allowing greater tax-free compounding, but to claim extra benefit when funds come from a tax-inefficient taxable account really takes credit for solving a problem which should have been addressed separately - i.e. improving the tax efficiency of the taxable account in the first place.

For an investor for whom muni's make sense and for whom equity in taxable are in index funds that shed minimal dividends and capital gains, this analysis makes little difference. The overwhelming factor in whether a conversion makes sense remains the investor's assumption that his/her marginal tax rate will be higher in the future. If the investor wants to act on this belief and convert, then paying taxes from non-tax deferred sources continues to be the way to go - cash flow needs permitting.
Munis have the the tax impact built into the yield (i.e., lower pre-tax return for the muni vs. a nominal bond). I don't understand why you consider a muni return in taxable a wash vs. a higher return on a regular bond in a Roth account. In today's low interest rate environment it might mot make much difference, but it could if interest rates increase back to historical yields.

User avatar
Cyclesafe
Posts: 658
Joined: Wed Dec 31, 2014 1:03 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by Cyclesafe » Tue Jul 24, 2018 10:06 am

cherijoh wrote:
Tue Jul 24, 2018 9:50 am
Cyclesafe wrote:
Tue Jul 24, 2018 9:38 am
I get how paying taxes on a conversion from a taxable account preserves tax-free space, allowing greater tax-free compounding, but to claim extra benefit when funds come from a tax-inefficient taxable account really takes credit for solving a problem which should have been addressed separately - i.e. improving the tax efficiency of the taxable account in the first place.

For an investor for whom muni's make sense and for whom equity in taxable are in index funds that shed minimal dividends and capital gains, this analysis makes little difference. The overwhelming factor in whether a conversion makes sense remains the investor's assumption that his/her marginal tax rate will be higher in the future. If the investor wants to act on this belief and convert, then paying taxes from non-tax deferred sources continues to be the way to go - cash flow needs permitting.
Munis have the the tax impact built into the yield (i.e., lower pre-tax return for the muni vs. a nominal bond). I don't understand why you consider a muni return in taxable a wash vs. a higher return on a regular bond in a Roth account. In today's low interest rate environment it might mot make much difference, but it could if interest rates increase back to historical yields.
That's not what I am saying. Of course, as you say, a muni will always(?) have a lower after tax return than a non-muni of roughly similar risk in a Roth. The Vanguard article presses the point that paying Roth conversion taxes separately out of taxable instead of from witholding from the converting t-IRA is especially beneficial when the taxable accounts are inefficient. I'm saying that the problem of tax inefficiency should have been addressed separately (with muni's or equity index funds) completely independently of the conversion decision.

retiringwhen
Posts: 456
Joined: Sat Jul 08, 2017 10:09 am
Location: New Jersey, USA

Re: Vanguard Paper on Roth Conversion Analysis

Post by retiringwhen » Tue Jul 24, 2018 10:36 am

Cyclesafe wrote:
Tue Jul 24, 2018 10:06 am
That's not what I am saying. Of course, as you say, a muni will always(?) have a lower after tax return than a non-muni of roughly similar risk in a Roth. The Vanguard article presses the point that paying Roth conversion taxes separately out of taxable instead of from witholding from the converting t-IRA is especially beneficial when the taxable accounts are inefficient. I'm saying that the problem of tax inefficiency should have been addressed separately (with muni's or equity index funds) completely independently of the conversion decision.
To be fair, the Vanguard papers shows the relative value of tax-efficient and tax-inefficient taxable funds as the source of paying taxes.

Both of them improve the BETR measurably but only difference is a matter of magnitude.

If someone needs to put Munis in their taxable account to increase their tax efficiency, then it seems they need to look for better places to put their money, the Roth makes a lot of sense. And, if break-even calculations make it look less desirable, then the investor is leaving a good opportunity off the table. This has been my case with a largish pre-tax IRA holding up back-door contributions in the future.

The tax improvement of using the taxable funds to pay the conversion is shown clearly to be tied to the length of expected investment horizon too. Something I had not considered myself in the past. The break-even only tax calculation essentially ignores this major opportunity cost of not making the conversion... Since many folks see Roth's as the last place to pull retirement funds, they by definition have the longest investment horizon, making this benefit probably much greater than initially considered.

This is a very thought provoking paper in a practical way for me. Time to go run some numbers.... (We had been doing small partial conversions the last couple of years, now I need to look at how high they can get to not go negative on the BETR to speed the availability of future back-door .)

cherijoh
Posts: 4945
Joined: Tue Feb 20, 2007 4:49 pm
Location: Charlotte NC

Re: Vanguard Paper on Roth Conversion Analysis

Post by cherijoh » Tue Jul 24, 2018 11:00 am

Cyclesafe wrote:
Tue Jul 24, 2018 10:06 am
cherijoh wrote:
Tue Jul 24, 2018 9:50 am
Cyclesafe wrote:
Tue Jul 24, 2018 9:38 am
I get how paying taxes on a conversion from a taxable account preserves tax-free space, allowing greater tax-free compounding, but to claim extra benefit when funds come from a tax-inefficient taxable account really takes credit for solving a problem which should have been addressed separately - i.e. improving the tax efficiency of the taxable account in the first place.

For an investor for whom muni's make sense and for whom equity in taxable are in index funds that shed minimal dividends and capital gains, this analysis makes little difference. The overwhelming factor in whether a conversion makes sense remains the investor's assumption that his/her marginal tax rate will be higher in the future. If the investor wants to act on this belief and convert, then paying taxes from non-tax deferred sources continues to be the way to go - cash flow needs permitting.
Munis have the the tax impact built into the yield (i.e., lower pre-tax return for the muni vs. a nominal bond). I don't understand why you consider a muni return in taxable a wash vs. a higher return on a regular bond in a Roth account. In today's low interest rate environment it might mot make much difference, but it could if interest rates increase back to historical yields.
That's not what I am saying. Of course, as you say, a muni will always(?) have a lower after tax return than a non-muni of roughly similar risk in a Roth. The Vanguard article presses the point that paying Roth conversion taxes separately out of taxable instead of from withholding from the converting t-IRA is especially beneficial when the taxable accounts are inefficient. I'm saying that the problem of tax inefficiency should have been addressed separately (with muni's or equity index funds) completely independently of the conversion decision.
I think you are underestimating the number of Boglehead converts who invest new money in tax-efficient index funds but have balked at paying the capital gains tax to convert all their existing taxable funds to indexes or ETFs in one fell swoop. Especially if their current actively-managed investments are in relatively low expense ratio funds - like VG's own actively managed funds.

Edited to add: I also think it depends on where you are in your investment journey. Relatively young investors have always had good exposure to index funds - older investors probably not so much. Especially since 401k contribution limits were quite a bit lower in the 80's and early 90s resulting in the need for more after-tax savings.

When I became a BH, I discontinued reinvesting capital gains and dividends in my active funds, but didn't sell all of them off for several years. (It would have bumped up my taxes a lot even with favorable cap gains rates). If I had seen this article after I rolled over my former employer's 401k into an IRA in 2004, it probably would have convinced me to do more extensive Roth conversions and get rid of my active funds sooner.
Last edited by cherijoh on Tue Jul 24, 2018 11:12 am, edited 1 time in total.

Nittany_Lion
Posts: 12
Joined: Fri Jul 20, 2018 10:49 am

Re: Vanguard Paper on Roth Conversion Analysis

Post by Nittany_Lion » Tue Jul 24, 2018 11:09 am

A quick question on how this works for folks who are already retired and thinking about a traditional-to-Roth conversion in an estate planning context.

Assuming the marginal tax rate of the heirs will be higher than that of their retiree parents -- and given all of the additional advantages outlined in the VG white paper -- it seems as though this will make sense for many seeking to leave a bequest (especially if federal tax rates trend up, as is widely expected to occur over the next 10-20 years).

The only potential rub would be the five-year holding period for penalty free withdrawals from the Roth. Does anyone know how that works for retirees?

Minot
Posts: 423
Joined: Sat Jan 29, 2011 2:35 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by Minot » Tue Jul 24, 2018 11:22 am

Eagle33 wrote:
Tue Jul 24, 2018 12:08 am
Miriam2 wrote:
Mon Jul 23, 2018 11:57 pm

Is there a link or window where we can find the Vanguard research or white papers?
link to Vanguard papers
https://investor.vanguard.com/investing ... t-research
The Filter on their Research Library page may also be useful.

retiringwhen
Posts: 456
Joined: Sat Jul 08, 2017 10:09 am
Location: New Jersey, USA

Re: Vanguard Paper on Roth Conversion Analysis

Post by retiringwhen » Tue Jul 24, 2018 11:25 am

Nittany_Lion wrote:
Tue Jul 24, 2018 11:09 am
A quick question on how this works for folks who are already retired and thinking about a traditional-to-Roth conversion in an estate planning context.

Assuming the marginal tax rate of the heirs will be higher than that of their retiree parents -- and given all of the additional advantages outlined in the VG white paper -- it seems as though this will make sense for many seeking to leave a bequest (especially if federal tax rates trend up, as is widely expected to occur over the next 10-20 years).

The only potential rub would be the five-year holding period for penalty free withdrawals from the Roth. Does anyone know how that works for retirees?
long post, but Michael Kitces covers all aspects of this. https://www.kitces.com/blog/understandi ... nversions/

TLDR version, If you are retired and over 59, then only the 5 year contribution rule would be in effect. Details in his article. There is a 5 year rule on distributions from a Roth IRA that says, regardless of age, 5 years must pass from the tax year of the initial Roth contribution/conversion before a penalty is applied to distributions of growth.

cherijoh
Posts: 4945
Joined: Tue Feb 20, 2007 4:49 pm
Location: Charlotte NC

Re: Vanguard Paper on Roth Conversion Analysis

Post by cherijoh » Tue Jul 24, 2018 11:33 am

Nittany_Lion wrote:
Tue Jul 24, 2018 11:09 am
A quick question on how this works for folks who are already retired and thinking about a traditional-to-Roth conversion in an estate planning context.

Assuming the marginal tax rate of the heirs will be higher than that of their retiree parents -- and given all of the additional advantages outlined in the VG white paper -- it seems as though this will make sense for many seeking to leave a bequest (especially if federal tax rates trend up, as is widely expected to occur over the next 10-20 years).

The only potential rub would be the five-year holding period for penalty free withdrawals from the Roth. Does anyone know how that works for retirees?
I think the problem is less significant if you've already hit the 5-year mark on a regular Roth account and are at least 59.5. This article might help.

GAAP
Posts: 586
Joined: Fri Apr 08, 2016 12:41 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by GAAP » Tue Jul 24, 2018 12:57 pm

And they completely ignore the option to pay those taxes with a HECM LOC -- zero cash-flow need for taxes, reduced portfolio withdrawals (since the taxes come from the LOC), improved conversion utilization (taxes from the LOC), higher net portfolio balance after conversion, etc.

retiringwhen
Posts: 456
Joined: Sat Jul 08, 2017 10:09 am
Location: New Jersey, USA

Re: Vanguard Paper on Roth Conversion Analysis

Post by retiringwhen » Tue Jul 24, 2018 1:27 pm

GAAP wrote:
Tue Jul 24, 2018 12:57 pm
And they completely ignore the option to pay those taxes with a HECM LOC -- zero cash-flow need for taxes, reduced portfolio withdrawals (since the taxes come from the LOC), improved conversion utilization (taxes from the LOC), higher net portfolio balance after conversion, etc.
HECM LOC as in Home Equity Line of Credit?

You suggesting using real estate leverage to boost your IRA account balances?

GAAP
Posts: 586
Joined: Fri Apr 08, 2016 12:41 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by GAAP » Tue Jul 24, 2018 2:05 pm

retiringwhen wrote:
Tue Jul 24, 2018 1:27 pm
HECM LOC as in Home Equity Line of Credit?

You suggesting using real estate leverage to boost your IRA account balances?
Yes -- and there's a fair amount of research to support that. Here are some starting points to consider before you criticize:
viewtopic.php?t=179189
https://www.thinkadvisor.com/2016/04/25 ... 0624145339
https://www.onefpa.org/journal/Pages/MA ... tgage.aspx
https://www.retirement-insight.com/use- ... etirement/

Keep in mind that equity in a house where you have no mortgage provides no direct value until you sell. A HECM LOC provides value from that equity with no negative impact to housing costs, cash flow, etc.

Yes, the HECM is due when you and your spouse die or move out -- but any balance would just reduce the sale proceeds. You still own the title to the house, not the bank -- you can't be evicted.

I was completely against reverse mortgages a year ago -- but now this is a core part of my strategy.

retiringwhen
Posts: 456
Joined: Sat Jul 08, 2017 10:09 am
Location: New Jersey, USA

Re: Vanguard Paper on Roth Conversion Analysis

Post by retiringwhen » Tue Jul 24, 2018 2:45 pm

GAAP wrote:
Tue Jul 24, 2018 2:05 pm
retiringwhen wrote:
Tue Jul 24, 2018 1:27 pm
HECM LOC as in Home Equity Line of Credit?

You suggesting using real estate leverage to boost your IRA account balances?
Yes -- and there's a fair amount of research to support that. Here are some starting points to consider before you criticize:
viewtopic.php?t=179189
https://www.thinkadvisor.com/2016/04/25 ... 0624145339
https://www.onefpa.org/journal/Pages/MA ... tgage.aspx
https://www.retirement-insight.com/use- ... etirement/

Keep in mind that equity in a house where you have no mortgage provides no direct value until you sell. A HECM LOC provides value from that equity with no negative impact to housing costs, cash flow, etc.

Yes, the HECM is due when you and your spouse die or move out -- but any balance would just reduce the sale proceeds. You still own the title to the house, not the bank -- you can't be evicted.

I was completely against reverse mortgages a year ago -- but now this is a core part of my strategy.
Okay, I didn't realize HECM is an acronym for a Reverse Mortgage. Frankly that term is new to me.....

Now I see how that could possibly work as a way to stretch tax-advantaged accounts.

I was thinking why would I take a out a HELOC with a 5-year repayment (or whatever) when I could pay the taxes from current assets

Miriam2
Posts: 2281
Joined: Fri Nov 14, 2014 11:51 am

Re: Vanguard Paper on Roth Conversion Analysis

Post by Miriam2 » Tue Jul 24, 2018 3:52 pm

Minot wrote:
Eagle33 wrote:
Miriam2 wrote: Is there a link or window where we can find the Vanguard research or white papers?
link to Vanguard papers
https://investor.vanguard.com/investing ... t-research
The Filter on their Research Library page may also be useful.
Thank you Eagle33 and Minot - those are VERY useful links :happy

I would try using the Vanguard search bar for articles I knew existed but could not find & the Vanguard search bar is not very good at pulling these articles up.

GAAP
Posts: 586
Joined: Fri Apr 08, 2016 12:41 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by GAAP » Tue Jul 24, 2018 4:08 pm

Miriam2 wrote:
Tue Jul 24, 2018 3:52 pm
Thank you Eagle33 and Minot - those are VERY useful links :happy

I would try using the Vanguard search bar for articles I knew existed but could not find & the Vanguard search bar is not very good at pulling these articles up.
The Vanguard web interface in general is not particularly good, I wouldn't feel bad about not finding something on that site...

User avatar
FiveK
Posts: 5567
Joined: Sun Mar 16, 2014 2:43 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FiveK » Tue Jul 24, 2018 8:17 pm

This is essentially what Maxing out your retirement accounts (Bogleheads wiki) discusses.

As noted there,
Numbers in these examples came from the 'Misc. calcs' tab (near row 150) in the personal finance toolbox spreadsheet. That spreadsheet can be used to plug in your own numbers.

Another spreadsheet that does similar calculations: https://docs.zoho.com/sheet/published.d ... 37d6fddfc7

jalbert
Posts: 3662
Joined: Fri Apr 10, 2015 12:29 am

Re: Vanguard Paper on Roth Conversion Analysis

Post by jalbert » Tue Jul 24, 2018 8:29 pm

Other factors can lower the BETR, e.g. LTC deductions, tax-loss harvesting in retirement, delaying SS and spending down a tIRA to replace the income likely reduces the tax drag as well in many cases.

I've been bothered by these analyses for some time, and I finally understand the concern well enough to articulate it. They use a deterministic marginal rate, when in fact lots of unforeseen circumstances and careful tax planning in retirement can have a significant effect on tax rates. Also, before RMDs, it is spending needs and not account balance that drives withdrawal rates and thus tax rates.
Risk is not a guarantor of return.

FinancialDave
Posts: 1579
Joined: Thu May 26, 2011 9:36 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FinancialDave » Wed Jul 25, 2018 12:49 am

jalbert wrote:
Tue Jul 24, 2018 8:29 pm
Other factors can lower the BETR, e.g. LTC deductions, tax-loss harvesting in retirement, delaying SS and spending down a tIRA to replace the income likely reduces the tax drag as well in many cases.

I've been bothered by these analyses for some time, and I finally understand the concern well enough to articulate it. They use a deterministic marginal rate, when in fact lots of unforeseen circumstances and careful tax planning in retirement can have a significant effect on tax rates. Also, before RMDs, it is spending needs and not account balance that drives withdrawal rates and thus tax rates.
You are definitely on to something here. I recently published an article (on a site I can't name) that shows there is much to this discussion beyond comparing two marginal tax rates.

During your working years it is generally true that your working marginal rate is the comparison line for later. Once you get to retirement it is quite normal that if you don't have a pension and you do have a large tIRA you will be spending down that tIRA in retirement across as many as 2-3 tax brackets, for many years. So, someone who put away money in the 30% bracket while working is really going to be at a great advantage to be able to spend that money in retirement, now in the 10%, 12%, 22%, & 24% tax brackets. If they need more income above that, they can certainly draw out some tax-free (Roth) funds, or almost tax-free taxable account funds. In such a case over a retirement starting at age 70, with a starting inflation adjusted income need of $100,000, having Roth or taxable account funds over 5% of the total, except as an emergency fund for larger expenses, is usually not necessary.

The other point that these studies do not address is the fact (at least from a historical perspective) that tax brackets and standard deductions have always had what I like to call "inflation creep" maybe not at actual inflation rates. In an article in 2015 I used 2% for this bracket creep based on history at the time. This year I used 1.5%, but over a 20 year retirement this still takes the top of the 24% bracket for a married couple from $315,000 to $418,000, so it is not trivial. It is in fact a bracket increase of 33%. I also inflated the standard deduction $100 per year. These inflationary aspects of our tax system are also items that favor the IRA over the Roth.

To tie all this back to the main Vanguard Paper I will suggest a few points:

1. Don't fall for the idea that converting your taxable account (and tIRA) into a Roth is going to get you some big bang for the effort. In most cases it will not, as most of the tax has already been paid on the front end of a taxable account and you probably don't need more Roth funds anyway. Just spend down the IRA funds instead, in what are probably lower tax brackets than when you were working, and then concentrate on having more efficient taxable funds that don't spin off a lot of dividends. Spend the taxable funds in the upper tax brackets where you might think about using Roth funds. Allocate the cost basis of the selling in an efficient manner to minimize increasing your tax bracket.

2. If your goal is getting rid of your IRA to be able to do a back door Roth, for many this is absolutely the wrong move to be making, at least with any sizable portion of your money. If you have a large traditional 401k then sure let's get rid of a small IRA, if it keeps you from filling out an 8606 for the rest of your life, if you need to do a Backdoor Roth.

3. The Vanguard article does highlight one important fact - once you do that Roth conversion (or contribution) your tax paid is locked in and you have lost all options (to save on taxes) except when to spend the tax-free money. If you keep that money in the tax-deferred status of an IRA you have options later to spend it across much lower tax brackets than when you were working.

4. Finally, as alluded to above, during your working years the measuring stick is your marginal tax rate but during retirement that measuring stick is the "effective tax rate" of the income need that is being pulled from the tax-deferred funds.

Dave
I love simulated data. It turns the impossible into the possible!

User avatar
FiveK
Posts: 5567
Joined: Sun Mar 16, 2014 2:43 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FiveK » Wed Jul 25, 2018 1:32 am

FinancialDave wrote:
Wed Jul 25, 2018 12:49 am
4. Finally, as alluded to above, during your working years the measuring stick is your marginal tax rate but during retirement that measuring stick is the "effective tax rate" of the income need that is being pulled from the tax-deferred funds.
Let's say someone can save 12% marginal now and expects retirement taxable income to come solely from traditional withdrawals. That person already has a large enough traditional balance to expect paying 22% marginal and 10% effective on withdrawals in retirement.

For any future contributions, should the person
- contribute to traditional because 12% > 10%, or
- contribute to Roth because 12% < 22%?

SGM
Posts: 2723
Joined: Wed Mar 23, 2011 4:46 am

Re: Vanguard Paper on Roth Conversion Analysis

Post by SGM » Wed Jul 25, 2018 4:53 am

Paying taxes for conversions was the only way I would have done a conversion. I recognized and have written about how there is a small advantage of decreasing ongoing tax drag on my taxable account which is lowered slightly by paying the taxes at the time of conversion. I don't think my taxable account was particularly tax inefficient and I use national munis, not state specific munis. I have converted all of my tax deferred accounts over a 5 year period when my income was a little lower due to going part time and then retiring completely. Also muni income increases my MAGI which increases my IRMMA premium.

By paying the taxes out of a taxable account I also do not lower the amount in tax advantaged space. I-orp predicted I would not need to tap my Roth accounts until many years from the day I made my first conversion. My Roth accounts have a lot of time ahead to grow without a withdrawal.

I will not have to take unneeded RMDs at age 70 1/2. With the rising markets recently both taxable and Roth accounts have already grown considerably. I was not anxious to see my taxable account going up with unneeded RMDs being added to it for the next 20 or more years.


The value of the Vanguard references for me is that they simply confirm my previous analysis of the value of Roth conversions and paying taxes out of a taxable account.

GAAP
Posts: 586
Joined: Fri Apr 08, 2016 12:41 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by GAAP » Wed Jul 25, 2018 9:15 am

The discussion around marginal rates assumes that there are no other taxes potentially due. In a state like the one I live in, there are also estate taxes to consider. If I convert to below the threshold, my heirs pay no estate taxes -- and those taxes would apply for both my death and my wife's death. The total estate value includes the amount of the TIRA that is actually due to the IRS. If I'm in the 22% bracket, then effectively 22% of the TIRA belongs to the IRS, yet would be subject to Washington's Estate tax.

FinancialDave
Posts: 1579
Joined: Thu May 26, 2011 9:36 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FinancialDave » Wed Jul 25, 2018 10:35 am

FiveK wrote:
Wed Jul 25, 2018 1:32 am
FinancialDave wrote:
Wed Jul 25, 2018 12:49 am
4. Finally, as alluded to above, during your working years the measuring stick is your marginal tax rate but during retirement that measuring stick is the "effective tax rate" of the income need that is being pulled from the tax-deferred funds.
Let's say someone can save 12% marginal now and expects retirement taxable income to come solely from traditional withdrawals. That person already has a large enough traditional balance to expect paying 22% marginal and 10% effective on withdrawals in retirement.

For any future contributions, should the person
- contribute to traditional because 12% > 10%, or
- contribute to Roth because 12% < 22%?
First let me premiss my answer by saying we never know when some "one-off" event that may spike our tax bracket for a year or three will come up.

But the answer to your direct question is you really need no more than about a 5% Roth to cover this scenerio effectively given the current tax structure we have and accounting for inflation for a starting retirement need in the low end of the 22% bracket. The reason is that even though you say you will be in the 22% marginal bracket, the majority of your IRA funds will be spent in the 12% to 0% tax brackets.

Dave
I love simulated data. It turns the impossible into the possible!

FinancialDave
Posts: 1579
Joined: Thu May 26, 2011 9:36 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FinancialDave » Wed Jul 25, 2018 10:54 am

SGM wrote:
Wed Jul 25, 2018 4:53 am
Paying taxes for conversions was the only way I would have done a conversion. I recognized and have written about how there is a small advantage of decreasing ongoing tax drag on my taxable account which is lowered slightly by paying the taxes at the time of conversion. I don't think my taxable account was particularly tax inefficient and I use national munis, not state specific munis. I have converted all of my tax deferred accounts over a 5 year period when my income was a little lower due to going part time and then retiring completely. Also muni income increases my MAGI which increases my IRMMA premium.

By paying the taxes out of a taxable account I also do not lower the amount in tax advantaged space. I-orp predicted I would not need to tap my Roth accounts until many years from the day I made my first conversion. My Roth accounts have a lot of time ahead to grow without a withdrawal.

I will not have to take unneeded RMDs at age 70 1/2. With the rising markets recently both taxable and Roth accounts have already grown considerably. I was not anxious to see my taxable account going up with unneeded RMDs being added to it for the next 20 or more years.


The value of the Vanguard references for me is that they simply confirm my previous analysis of the value of Roth conversions and paying taxes out of a taxable account.
First I failed to mention that the only way to pay the tax if you are less than 59.5 is of course to use a taxable account or other savings.

The fact that you had a few low income years to do the conversions is also a good thing, BUT the time in the Roth is not a factor in the equation, only the tax difference. Sure if the account grows in the Roth that is good, but the funds left in the IRA would have grown by the same amount and your only gain (in spendable income) or loss is due to tax differencial. This is the part many people get wrong and in many cases it harms them because once they do the conversion their taxes are locked in. In your case you picked what seemed like the best time to do the conversion, so if it turns out the conversion can be utilized in a higher tax bracket it will have been a good choice.

Finally, for your bolded text above about Vanguard, I don't think they would be so bold as to know if it makes sense for you to do a Roth Conversion, only that if you are going to do it there is some minor gain to use taxable funds to pay the tax.

One thing to remember is that a taxable account takes on much of the characteristic of a Roth account to begin with, so making your taxable account smaller and your Roth larger will not be helpful if you run out of tIRA funds later in your retirement.

In retirement tax-deferred funds should be spent down first in the lower tax brackets and then taxable and last Roth funds used to keep you out of the higher tax brackets.

Dave
I love simulated data. It turns the impossible into the possible!

User avatar
FiveK
Posts: 5567
Joined: Sun Mar 16, 2014 2:43 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FiveK » Wed Jul 25, 2018 10:55 am

FinancialDave wrote:
Wed Jul 25, 2018 10:35 am
FiveK wrote:
Wed Jul 25, 2018 1:32 am
FinancialDave wrote:
Wed Jul 25, 2018 12:49 am
4. Finally, as alluded to above, during your working years the measuring stick is your marginal tax rate but during retirement that measuring stick is the "effective tax rate" of the income need that is being pulled from the tax-deferred funds.
Let's say someone can save 12% marginal now and expects retirement taxable income to come solely from traditional withdrawals. That person already has a large enough traditional balance to expect paying 22% marginal and 10% effective on withdrawals in retirement.

For any future contributions, should the person
- contribute to traditional because 12% > 10%, or
- contribute to Roth because 12% < 22%?
But the answer to your direct question is you really need no more than about a 5% Roth to cover this scenerio effectively given the current tax structure we have and accounting for inflation for a starting retirement need in the low end of the 22% bracket. The reason is that even though you say you will be in the 22% marginal bracket, the majority of your IRA funds will be spent in the 12% to 0% tax brackets.
That most of the withdrawal will be taxed at 0-12% is correct, but also irrelevant. That's because those withdrawals are based on money contributed in the past.

In the given situation, even if no further traditional contributions are made, the person is already paying 22% on some amount. Further contributions will only increase the amount withdrawn and incurring a 22% tax.

Thus the effective rate is irrelevant: new contributions are withdrawn "on top of" withdrawals due to past contributions, so the destination of any new contribution should not be judged by the effective withdrawal rate, but by the withdrawal marginal tax rate.

See Traditional versus Roth - Bogleheads and Marginal Vs Effective Tax Rates And When To Use Each for more.
Last edited by FiveK on Wed Jul 25, 2018 11:07 am, edited 1 time in total.

GAAP
Posts: 586
Joined: Fri Apr 08, 2016 12:41 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by GAAP » Wed Jul 25, 2018 10:58 am

Also note, as the tax law is currently written, we go back to the old brackets in 2025. The 22% bracket today becomes 25% in 7 years -- making conversions up to the top of the 24% bracket potentially of value. The same concept applies for the other brackets.

FinancialDave
Posts: 1579
Joined: Thu May 26, 2011 9:36 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FinancialDave » Wed Jul 25, 2018 11:34 am

GAAP wrote:
Wed Jul 25, 2018 9:15 am
The discussion around marginal rates assumes that there are no other taxes potentially due. In a state like the one I live in, there are also estate taxes to consider. If I convert to below the threshold, my heirs pay no estate taxes -- and those taxes would apply for both my death and my wife's death. The total estate value includes the amount of the TIRA that is actually due to the IRS. If I'm in the 22% bracket, then effectively 22% of the TIRA belongs to the IRS, yet would be subject to Washington's Estate tax.
Don't take this as Estate planning advice, but do seek the proper estate planning advice if your estate is over $2 million.

I am not exactly sure what you are saying as your Federal marginal tax rates have little to do with the Washington State Estate tax rates.

The exclusion amount before any estate tax is $2,193,000. So if one spouse dies with a $4 million estate, their half of the estate is only $2 million, so no tax is due. Now if the estate is $6,000,000 when the second spouse dies then the tax is roughly 10% effective tax on the $6m.

So does it seem reasonable to spend 22% to convert tIRA funds to avoid zero state (and federal) tax on the first spouse's death and 9% on the second's?

Check with your estate attorney.
I love simulated data. It turns the impossible into the possible!

GAAP
Posts: 586
Joined: Fri Apr 08, 2016 12:41 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by GAAP » Wed Jul 25, 2018 12:12 pm

FinancialDave wrote:
Wed Jul 25, 2018 11:34 am

I am not exactly sure what you are saying as your Federal marginal tax rates have little to do with the Washington State Estate tax rates.
I didn't say they did.

If you own a TIRA, a portion of that TIRA is effectively owned by the IRS -- they will get their money when you withdraw it, or when your heirs do. That portion is also potentially subject to Washington estate tax. When you do a Roth conversion, that portion is paid to the IRS, but also removed from your estate and no longer subject to the Washington estate tax.

Yes there's an exclusion amount. If your estate is below that limit this doesn't matter. If your estate is above that limit, then it does matter. A Roth conversion can actually take you from "it matters" to "it doesn't matter". The limit is specific to the state of residence, and the whole concept only applies to states with an estate tax.

eli80
Posts: 25
Joined: Thu Dec 08, 2016 10:49 am

Re: Vanguard Paper on Roth Conversion Analysis

Post by eli80 » Wed Jul 25, 2018 12:20 pm

FinancialDave wrote:
Wed Jul 25, 2018 12:49 am
jalbert wrote:
Tue Jul 24, 2018 8:29 pm
Other factors can lower the BETR, e.g. LTC deductions, tax-loss harvesting in retirement, delaying SS and spending down a tIRA to replace the income likely reduces the tax drag as well in many cases.

I've been bothered by these analyses for some time, and I finally understand the concern well enough to articulate it. They use a deterministic marginal rate, when in fact lots of unforeseen circumstances and careful tax planning in retirement can have a significant effect on tax rates. Also, before RMDs, it is spending needs and not account balance that drives withdrawal rates and thus tax rates.
You are definitely on to something here. I recently published an article (on a site I can't name) that shows there is much to this discussion beyond comparing two marginal tax rates.

During your working years it is generally true that your working marginal rate is the comparison line for later. Once you get to retirement it is quite normal that if you don't have a pension and you do have a large tIRA you will be spending down that tIRA in retirement across as many as 2-3 tax brackets, for many years. So, someone who put away money in the 30% bracket while working is really going to be at a great advantage to be able to spend that money in retirement, now in the 10%, 12%, 22%, & 24% tax brackets. If they need more income above that, they can certainly draw out some tax-free (Roth) funds, or almost tax-free taxable account funds. In such a case over a retirement starting at age 70, with a starting inflation adjusted income need of $100,000, having Roth or taxable account funds over 5% of the total, except as an emergency fund for larger expenses, is usually not necessary.

The other point that these studies do not address is the fact (at least from a historical perspective) that tax brackets and standard deductions have always had what I like to call "inflation creep" maybe not at actual inflation rates. In an article in 2015 I used 2% for this bracket creep based on history at the time. This year I used 1.5%, but over a 20 year retirement this still takes the top of the 24% bracket for a married couple from $315,000 to $418,000, so it is not trivial. It is in fact a bracket increase of 33%. I also inflated the standard deduction $100 per year. These inflationary aspects of our tax system are also items that favor the IRA over the Roth.

To tie all this back to the main Vanguard Paper I will suggest a few points:

1. Don't fall for the idea that converting your taxable account (and tIRA) into a Roth is going to get you some big bang for the effort. In most cases it will not, as most of the tax has already been paid on the front end of a taxable account and you probably don't need more Roth funds anyway. Just spend down the IRA funds instead, in what are probably lower tax brackets than when you were working, and then concentrate on having more efficient taxable funds that don't spin off a lot of dividends. Spend the taxable funds in the upper tax brackets where you might think about using Roth funds. Allocate the cost basis of the selling in an efficient manner to minimize increasing your tax bracket.

2. If your goal is getting rid of your IRA to be able to do a back door Roth, for many this is absolutely the wrong move to be making, at least with any sizable portion of your money. If you have a large traditional 401k then sure let's get rid of a small IRA, if it keeps you from filling out an 8606 for the rest of your life, if you need to do a Backdoor Roth.

3. The Vanguard article does highlight one important fact - once you do that Roth conversion (or contribution) your tax paid is locked in and you have lost all options (to save on taxes) except when to spend the tax-free money. If you keep that money in the tax-deferred status of an IRA you have options later to spend it across much lower tax brackets than when you were working.

4. Finally, as alluded to above, during your working years the measuring stick is your marginal tax rate but during retirement that measuring stick is the "effective tax rate" of the income need that is being pulled from the tax-deferred funds.

Dave
FinancialDave - I have done much reading and running numbers in i-orp and other calculators regarding Roth conversions. I think my conclusion jives with what you are saying. My conclusion was that for the individual considering the conversions, Roth conversions don't really have a significant impact on the ability to draw down one's accounts. That is, the individual may not necessarily be able to spend more due to Roth conversions. However, they may leave their heirs in better shape. So if you plan on spending your last dollar when you take your last breath, there is no reason to do Roth conversions. Do you think that is a fair conclusion?

Thanks.

FinancialDave
Posts: 1579
Joined: Thu May 26, 2011 9:36 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FinancialDave » Wed Jul 25, 2018 2:18 pm

GAAP wrote:
Wed Jul 25, 2018 12:12 pm
FinancialDave wrote:
Wed Jul 25, 2018 11:34 am

I am not exactly sure what you are saying as your Federal marginal tax rates have little to do with the Washington State Estate tax rates.
I didn't say they did.

If you own a TIRA, a portion of that TIRA is effectively owned by the IRS -- they will get their money when you withdraw it, or when your heirs do. That portion is also potentially subject to Washington estate tax. When you do a Roth conversion, that portion is paid to the IRS, but also removed from your estate and no longer subject to the Washington estate tax.

Yes there's an exclusion amount. If your estate is below that limit this doesn't matter. If your estate is above that limit, then it does matter. A Roth conversion can actually take you from "it matters" to "it doesn't matter". The limit is specific to the state of residence, and the whole concept only applies to states with an estate tax.
Everything you say is true, but I think you missed the point that just because you moved future tax to "it doesn't matter" does not mean you and your heirs have more spendable income. Paying 22% now is not always beneficial to paying zero later, especially if the alternative is paying zero tax now and 9% later. Just because an IRA is part of an estate does not make the Federal tax due on it all at once and in the case of a spouse it is all rolled into their account and follows their RMD age.
I love simulated data. It turns the impossible into the possible!

FinancialDave
Posts: 1579
Joined: Thu May 26, 2011 9:36 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FinancialDave » Wed Jul 25, 2018 2:22 pm

GAAP wrote:
Wed Jul 25, 2018 10:58 am
Also note, as the tax law is currently written, we go back to the old brackets in 2025. The 22% bracket today becomes 25% in 7 years -- making conversions up to the top of the 24% bracket potentially of value. The same concept applies for the other brackets.
Ya, there were a lot of really smart people saying the Bush tax cuts would expire as well and we all can see how that turned out.

I am not saying it won't happen, I just don't base my decisions on what "could" happen, I base them on the current facts, and I attach a small risk premium on having a diversified strategy.
I love simulated data. It turns the impossible into the possible!

FinancialDave
Posts: 1579
Joined: Thu May 26, 2011 9:36 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FinancialDave » Wed Jul 25, 2018 2:36 pm

FiveK wrote:
Wed Jul 25, 2018 10:55 am
FinancialDave wrote:
Wed Jul 25, 2018 10:35 am
FiveK wrote:
Wed Jul 25, 2018 1:32 am
FinancialDave wrote:
Wed Jul 25, 2018 12:49 am
4. Finally, as alluded to above, during your working years the measuring stick is your marginal tax rate but during retirement that measuring stick is the "effective tax rate" of the income need that is being pulled from the tax-deferred funds.
Let's say someone can save 12% marginal now and expects retirement taxable income to come solely from traditional withdrawals. That person already has a large enough traditional balance to expect paying 22% marginal and 10% effective on withdrawals in retirement.

For any future contributions, should the person
- contribute to traditional because 12% > 10%, or
- contribute to Roth because 12% < 22%?
But the answer to your direct question is you really need no more than about a 5% Roth to cover this scenerio effectively given the current tax structure we have and accounting for inflation for a starting retirement need in the low end of the 22% bracket. The reason is that even though you say you will be in the 22% marginal bracket, the majority of your IRA funds will be spent in the 12% to 0% tax brackets.
That most of the withdrawal will be taxed at 0-12% is correct, but also irrelevant. That's because those withdrawals are based on money contributed in the past.

In the given situation, even if no further traditional contributions are made, the person is already paying 22% on some amount. Further contributions will only increase the amount withdrawn and incurring a 22% tax.

Thus the effective rate is irrelevant: new contributions are withdrawn "on top of" withdrawals due to past contributions, so the destination of any new contribution should not be judged by the effective withdrawal rate, but by the withdrawal marginal tax rate.

See Traditional versus Roth - Bogleheads and Marginal Vs Effective Tax Rates And When To Use Each for more.
I think you missed my point - if you have at least 5% of your savings going toward the Roth (95% IRA / 5% Roth) then your retirement tax bracket never goes above 12%, because you spend down the Roth money only when necessary.

It is certainly not irrelevant that you are spending 95% of your money in tax brackets 0 to 12% and then the other 5% comes from your Roth.

If you run out of tIRA funds in retirement under your scenerio, what happens -- now you are spending VERY expensive Roth money (12% tax already paid) in the ZERO percent tax bracket. Granted this is not as bad as most of us that paid 25% or greater tax on Roth contributions or conversions, but it still will NOT leave you with the most after-tax spendable income.

In your example there is NO 22% tax bracket in retirement, if you have even a small amount of Roth funds!

dave
I love simulated data. It turns the impossible into the possible!

FinancialDave
Posts: 1579
Joined: Thu May 26, 2011 9:36 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FinancialDave » Wed Jul 25, 2018 2:52 pm

eli80 wrote:
Wed Jul 25, 2018 12:20 pm
FinancialDave - I have done much reading and running numbers in i-orp and other calculators regarding Roth conversions. I think my conclusion jives with what you are saying. My conclusion was that for the individual considering the conversions, Roth conversions don't really have a significant impact on the ability to draw down one's accounts. That is, the individual may not necessarily be able to spend more due to Roth conversions. However, they may leave their heirs in better shape. So if you plan on spending your last dollar when you take your last breath, there is no reason to do Roth conversions. Do you think that is a fair conclusion?

Thanks.
I would answer this question with another question. If you have studied the math you will know that more spendable income comes from the difference in your average Roth contribution and conversion tax level compared to whoever spends the funds and their tax bracket. In the case of heirs this amounts to what decisions you have made on when to pay taxes on Roth contributions and conversions vs. when your heirs would have to pay taxes on any tIRA funds they receive and the tax rate they would pay.

What essentially happens is the same thing that happens to you, only just the names have changed, EXCEPT that if your heirs are young with children some tax credits could be lost if their inheritance is too large and its all in an IRA.

In general I doubt your heirs are going to be in a higher tax bracket than you (when you were working) so I don't think the math is all that different. Of course people love to get tax-free money, but that is not the question here - the question is in which case do you have more spendable income and that is when you understand the rules (which I think you do) and don't stray too far from them.
I love simulated data. It turns the impossible into the possible!

User avatar
FiveK
Posts: 5567
Joined: Sun Mar 16, 2014 2:43 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FiveK » Wed Jul 25, 2018 3:02 pm

FinancialDave wrote:
Wed Jul 25, 2018 2:36 pm
I think you missed my point - if you have at least 5% of your savings going toward the Roth (95% IRA / 5% Roth) then your retirement tax bracket never goes above 12%, because you spend down the Roth money only when necessary.
Sigh.... The low bandwidth of forum posting....

Perhaps you have another constraint in mind? For example, someone with $1.9 million in traditional and $100K in Roth is definitely in the 22% bracket with a 4% withdrawal rate. Unless of course the withdrawal is $80K from the Roth, but that merely postpones the inevitable. ;)
It is certainly not irrelevant that you are spending 95% of your money in tax brackets 0 to 12% and then the other 5% comes from your Roth.
The relevance question is about the tax measurement: effective vs. marginal. If you are saying that one should look at marginal retirement rates to evaluate traditional vs. Roth contributions then we agree.
If you run out of tIRA funds in retirement under your scenerio, what happens -- now you are spending VERY expensive Roth money (12% tax already paid) in the ZERO percent tax bracket. Granted this is not as bad as most of us that paid 25% or greater tax on Roth contributions or conversions, but it still will NOT leave you with the most after-tax spendable income.
Agreed. It is certainly possible to have too much in Roth/traditional and therefore not enough in traditional/Roth.
In your example there is NO 22% tax bracket in retirement, if you have even a small amount of Roth funds!
Hey, it's my example (aka, this round was on me) so let's go with it as stated, ok? :beer

The example stipulated that the person already expected to pay 22% on some amount of withdrawal and asked whether that 22%, or the 10% effective tax, was the relevant rate for evaluating a new contribution. Which do you think is the relevant withdrawal rate, 10% or 22%, for this new contribution?

FinancialDave
Posts: 1579
Joined: Thu May 26, 2011 9:36 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FinancialDave » Wed Jul 25, 2018 4:38 pm

FiveK wrote:
Wed Jul 25, 2018 3:02 pm


Perhaps you have another constraint in mind? For example, someone with $1.9 million in traditional and $100K in Roth is definitely in the 22% bracket with a 4% withdrawal rate. Unless of course the withdrawal is $80K from the Roth, but that merely postpones the inevitable. ;)


The relevance question is about the tax measurement: effective vs. marginal. If you are saying that one should look at marginal retirement rates to evaluate traditional vs. Roth contributions then we agree.


Hey, it's my example (aka, this round was on me) so let's go with it as stated, ok? :beer

The example stipulated that the person already expected to pay 22% on some amount of withdrawal and asked whether that 22%, or the 10% effective tax, was the relevant rate for evaluating a new contribution. Which do you think is the relevant withdrawal rate, 10% or 22%, for this new contribution?
Yes, low bandwidth.

Sorry,
My constraint was for a reasonable couple who had saved a combined pre-tax amount of $1million split ($980k IRA $20K Roth). My assumption was they knew when to spend the Roth and when to spend the IRA.

I will try your example as stated, but I know nothing about how much Roth is saved so far and at what tax rate -- that is not irrelevant.

Let me state it a different way. When all is said and done the optimal Roth / tIRA choice is defined by the AVERAGE tax rate paid on all your Roth contributions and conversions, unless of course you think you are going to "order" each one of them and spend them accordingly on the other end. That must be weighed against all tax brackets in which the retirement savings are spent, not just the marginal one! This is not really debatable though many think it is.

To be frank, it makes absolutely no difference what your working tax rate is, the only thing that matters prior to retirement is the tax paid on the Roth funds.

Let me try and get to the point of what you are asking, given the following (to improve the bandwidth):

1. You have not saved any Roth funds - secondly you will only save Roth funds if your tax rate is 12%.

2. You have not saved any IRA funds - you will save them in all marginal tax rates as it makes no difference to your retirement, as long as you save enough - which I will define shortly. Of course the Roth / IRA split is controlled by how much pre-tax income you can afford to save and the size of the buckets that are available.

3. Given you "think" you will need to spend money in the 22% bracket, then you NEED some Roth funds saved in your current retirement plan - the question is how much?? I still ask you though, how much of the Roth funds are you going to use in the 22% bracket - are you merely keeping yourself in the 22% bracket and avoiding the 24% bracket, or do you plan to use the Roth to pull you down to the 12% bracket?

4. Because you also stated (if I remember correctly) that all your retirement spending is coming from your retirement accounts - no SS or pension, that is the second reason why your scenario highly favors the larger IRA balance.

I think I answered your question as to your 22% MARGINAL rate is a line that tells you that you need SOME Roth money but it doesn't tell you how much. :oops:

Let's look at a single person and how much money they will spend in this case BELOW the 22% bracket:

1. Above age 65 $13,600 0% tax
2. next $9725 income $953 tax
3. between $9525 to $38,700 is $3,501 tax

So to answer your question properly you have to know "effectively" how much money you are spending in retirement from your tax advantaged (and taxable account because it is much like a Roth).

Your marginal rate tells you if you need any Roth at all, your effective tax from these accounts tells you how much.

In the above example you should be spending $52,300 from your IRA and the rest from your Roth to stay on the edge of the 22% bracket. Now if you have SS you need to work that into the mix. The current size of your Roth / traditional funds is technically what determines whether more money should be going towards the Roth. Given the conditions you state all I know is you need some Roth - it could be 3% or it could be 25%, but it will be as I mentioned, based on your average Roth contribution tax paid and your retirement tax profile for all your pre-tax and post-tax funds spent.

dave
I love simulated data. It turns the impossible into the possible!

BigJohn
Posts: 1581
Joined: Wed Apr 02, 2014 11:27 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by BigJohn » Wed Jul 25, 2018 5:32 pm

FinancialDave wrote:
Wed Jul 25, 2018 2:36 pm
I think you missed my point - if you have at least 5% of your savings going toward the Roth (95% IRA / 5% Roth) then your retirement tax bracket never goes above 12%, because you spend down the Roth money only when necessary.
I’m missing something here when you say “never” goes above 12%. It doesn’t seem too difficult to get your retirement tax bracket above that once RMDs start. If you rolled over your 401k into an IRA and took your pension as a lump sum as well you likely have a fairly sizable tIRA balance. Let’s say it’s $2M when you start RMDs. In the first year the RMD is $73K which is well into the 22% bracket for s single filer and close for MFJ filers.

I’m not saying your analysis isn’t valid for some but I don’t see how it can be valid for all. What am I missing?

GAAP
Posts: 586
Joined: Fri Apr 08, 2016 12:41 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by GAAP » Wed Jul 25, 2018 5:46 pm

BigJohn wrote:
Wed Jul 25, 2018 5:32 pm
It doesn’t seem too difficult to get your retirement tax bracket above that once RMDs start.
Especially after the first spouse dies and the second is forced into the single tax brackets...

FinancialDave
Posts: 1579
Joined: Thu May 26, 2011 9:36 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FinancialDave » Wed Jul 25, 2018 6:08 pm

BigJohn wrote:
Wed Jul 25, 2018 5:32 pm
FinancialDave wrote:
Wed Jul 25, 2018 2:36 pm
I think you missed my point - if you have at least 5% of your savings going toward the Roth (95% IRA / 5% Roth) then your retirement tax bracket never goes above 12%, because you spend down the Roth money only when necessary.
I’m missing something here when you say “never” goes above 12%. It doesn’t seem too difficult to get your retirement tax bracket above that once RMDs start. If you rolled over your 401k into an IRA and took your pension as a lump sum as well you likely have a fairly sizable tIRA balance. Let’s say it’s $2M when you start RMDs. In the first year the RMD is $73K which is well into the 22% bracket for s single filer and close for MFJ filers.

I’m not saying your analysis isn’t valid for some but I don’t see how it can be valid for all. What am I missing?
Ok, never was not to imply the 1% can't have enough tax deferred money to get them into the 22% bracket, but it's not likely they did it while working in the 12% bracket their whole life, so I am just trying to be reasonable for the example stated. Once again you have to look at how much of that $73k is spent in the 22% and higher tax bracket and what is the tax bracket of your Roth funds -- for me that average tax bracket is > 25% and I don't even have $1m in my IRA.

I'll agree "never" is a poor choice. It all has to do with tax differentials. As your average Roth tax bracket goes up to 25-30%, the problem is the same only now you are trying to stay in the 24% tax bracket, which even for a single person is north of $170,000.

Dave
I love simulated data. It turns the impossible into the possible!

FinancialDave
Posts: 1579
Joined: Thu May 26, 2011 9:36 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FinancialDave » Wed Jul 25, 2018 7:03 pm

BigJohn wrote:
Wed Jul 25, 2018 5:32 pm
FinancialDave wrote:
Wed Jul 25, 2018 2:36 pm
I think you missed my point - if you have at least 5% of your savings going toward the Roth (95% IRA / 5% Roth) then your retirement tax bracket never goes above 12%, because you spend down the Roth money only when necessary.
I’m missing something here when you say “never” goes above 12%. It doesn’t seem too difficult to get your retirement tax bracket above that once RMDs start. If you rolled over your 401k into an IRA and took your pension as a lump sum as well you likely have a fairly sizable tIRA balance. Let’s say it’s $2M when you start RMDs. In the first year the RMD is $73K which is well into the 22% bracket for s single filer and close for MFJ filers.

I’m not saying your analysis isn’t valid for some but I don’t see how it can be valid for all. What am I missing?
In the case of someone who took their pension as a lump sum (ugh), I hope they did it early enough to convert $900k of it to a Roth (leaving them $675k -- converted at average 25% bracket), to match up with now a smaller $1.1 million in the IRA. I also hope they stay married because being single is the "pits" in this case.

If they are lucky, this person with $2 million of pre-tax saved in the right proportions (1.1 m / $675k) IRA/Roth can certainly stay in the 12% tax bracket for at least 19 years starting at age 70. I assume 1.5% inflation "creep" in tax brackets and $100 per year in standard deduction increase for the married couple. I also assume retirement growth rate of funds at 6%, starting income need of $150k inflated at 1.5% per year.

Dave
Last edited by FinancialDave on Wed Jul 25, 2018 7:13 pm, edited 2 times in total.
I love simulated data. It turns the impossible into the possible!

randomguy
Posts: 6284
Joined: Wed Sep 17, 2014 9:00 am

Re: Vanguard Paper on Roth Conversion Analysis

Post by randomguy » Wed Jul 25, 2018 7:07 pm

jalbert wrote:
Tue Jul 24, 2018 8:29 pm
Other factors can lower the BETR, e.g. LTC deductions, tax-loss harvesting in retirement, delaying SS and spending down a tIRA to replace the income likely reduces the tax drag as well in many cases.

I've been bothered by these analyses for some time, and I finally understand the concern well enough to articulate it. They use a deterministic marginal rate, when in fact lots of unforeseen circumstances and careful tax planning in retirement can have a significant effect on tax rates. Also, before RMDs, it is spending needs and not account balance that drives withdrawal rates and thus tax rates.
They are making assumptions which in this case are basically measuring tax drag. As your tax drag goes to zero (say your LTGC rate is 0, you make charitable contributions instead of selling, stepped up cost basis on death,...) the gap will shrink.

And yes to a large extent the unknown dominates. You really want to look at a range of numbers. It might turn out that the numbers should be something like 35%+-10% and 29.6%+-10%. On average you will win by picking the 29.6% case. But it might turn out in the worst case the numbers switch to 25% versus 35% instead of 35% vs 29.6%. You then have to balance worst case versus best case and pick what works for you.

User avatar
FiveK
Posts: 5567
Joined: Sun Mar 16, 2014 2:43 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FiveK » Wed Jul 25, 2018 7:13 pm

FinancialDave wrote:
Wed Jul 25, 2018 4:38 pm
Let me state it a different way. When all is said and done the optimal Roth / tIRA choice....
...
Your marginal rate tells you if you need any Roth at all, your effective tax from these accounts tells you how much.
...
The current size of your Roth / traditional funds is technically what determines whether more money should be going towards the Roth.
Before continuing point by point discussion, it may be that we have one "side" saying the sky is blue while the counterargument is that the grass is green.

From a quick read I can agree with the items above, from the perspective that (I think) you are looking at the split between traditional and Roth over an entire career. On the other hand, the wiki and Kitces articles, etc., are looking at how one might choose for a specific year's contribution.

Does that seem correct to you?

jalbert
Posts: 3662
Joined: Fri Apr 10, 2015 12:29 am

Re: Vanguard Paper on Roth Conversion Analysis

Post by jalbert » Wed Jul 25, 2018 7:18 pm

Perhaps you have another constraint in mind? For example, someone with $1.9 million in traditional and $100K in Roth is definitely in the 22% bracket with a 4% withdrawal rate. Unless of course the withdrawal is $80K from the Roth, but that merely postpones the inevitable
If a 4% withdrawal rate on that tIRA balance meets that person's expenses, why would they withdraw more from the tIRA if the balance were higher? Unless they had to meet RMDs' wouldn't they lower the withdrawal percentage so that they only withdrew what they needed to spend?

Thus, you can't even begin to develop an accurate model without separating withdrawal periods to the time before RMDs and the time after RMDs.

And because events that affect future tax rates are unpredictable, not just tax code changes, but personal circumstance, the best analysis would not confuse strategy and outcome by using a specific marginal tax rate that you think is most likely, but would use the probabilistic expected value of the marginal tax rate viewed as a random variable.

For instance, if you plan to pay out of pocket for long-term care, either in full or until running out of funds and going on medicaid, that is a large medical deduction that will greatly extend the amount of a tIRA that will be withdrawn tax free. In an ideal analysis you would include those outcomes weighted by their probability of happening in calculating the amount to convert. If in the end, you did not need long-term care, it would be confusing outcome with strategy to infer that more tIRA funds should have been converted.
Risk is not a guarantor of return.

User avatar
One Ping
Posts: 481
Joined: Thu Sep 24, 2015 4:53 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by One Ping » Wed Jul 25, 2018 7:19 pm

GAAP wrote:
Wed Jul 25, 2018 5:46 pm
BigJohn wrote:
Wed Jul 25, 2018 5:32 pm
It doesn’t seem too difficult to get your retirement tax bracket above that once RMDs start.
Especially after the first spouse dies and the second is forced into the single tax brackets...
Yeah, that single tax bracket is a killer. That's why we are converting into the 22% bracket. Survivor will definitely be in the 22% single bracket, if not the 24%. I'm not sure people realize the impact of being single on their tax bracket after being MFJ for a while. It's something to consider ...
"Re-verify our range to target ... one ping only."

BigJohn
Posts: 1581
Joined: Wed Apr 02, 2014 11:27 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by BigJohn » Wed Jul 25, 2018 7:32 pm

FinancialDave wrote:
Wed Jul 25, 2018 7:03 pm
In the case of someone who took their pension as a lump sum (ugh), I hope they did it early enough to convert $900k of it to a Roth (leaving them $675k -- converted at average 25% bracket), to match up with now a smaller $1.1 million in the IRA. I also hope they stay married because being single is the "pits" in this case.

If they are lucky, this person with $2 million of pre-tax saved in the right proportions (1.1 m / $675k) IRA/Roth can certainly stay in the 12% tax bracket for at least 19 years starting at age 70. I assume 1.5% inflation "creep" in tax brackets and $100 per year in standard deduction increase for the married couple. I also assume retirement growth rate of funds at 6%, starting income need of $150k inflated at 1.5% per year.
Dave, just FYI...

I took my pension as a lump sum because it was available under grandfathered provisions tied to 95% of the 30 year treasury at the time when it was at it's lowest in 30+ years. Just too good a deal to pass up plus.... it keeps my yearly income lower and more under my control to facilitate Roth conversions. Taking the pension would have significantly reduced my tax efficient Roth conversion window.

My current plan has me near 50/50 tIRA/Roth at age 70 (in 8 years) when both RMDs and SS both kick-in (the dreaded tax torpedo). This plan is certainly not optimum since that can only be known in hindsight due to uncertainty in longevity, return rates, future tax law changes, etc. However, I know in which scenarios I win and in which I lose and am fully prepared to accept the inherent risk.

Staying married is not always a choice. Death doesn't really care about your tax plans so I'm filing single for the first time this year.

All this just to say..... there are a lot of individual circumstances in these decisions and it's never a one-size-fits-all solution.

User avatar
FiveK
Posts: 5567
Joined: Sun Mar 16, 2014 2:43 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FiveK » Wed Jul 25, 2018 7:35 pm

jalbert wrote:
Wed Jul 25, 2018 7:18 pm
Perhaps you have another constraint in mind? For example, someone with $1.9 million in traditional and $100K in Roth is definitely in the 22% bracket with a 4% withdrawal rate. Unless of course the withdrawal is $80K from the Roth, but that merely postpones the inevitable
If a 4% withdrawal rate on that tIRA balance meets that person's expenses, why would they withdraw more from the tIRA if the balance were higher? Unless they had to meet RMDs' wouldn't they lower the withdrawal percentage so that they only withdrew what they needed to spend?
...and if money remains unspent at death, one would have to know the marginal rates applicable to the heirs, and how much was left to each, etc., etc. That gets messy very quickly, adding more noise than signal.

In one extreme, someone with no heirs other than charities could indeed assume a marginal withdrawal tax rate of 0% once some minimum invested balance was reached.

At the other extreme, someone could plan to blow the entire "excess" traditional balance in one shot (say, a Lamborghini, a couple of Ferraris, and some Teslas would be nice) and the expected marginal withdrawal tax rate is then at maximum.

In between, assuming that spending becomes a direct function of total spendable makes things tractable and is not an unreasonable model.

BigJohn
Posts: 1581
Joined: Wed Apr 02, 2014 11:27 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by BigJohn » Wed Jul 25, 2018 7:38 pm

One Ping wrote:
Wed Jul 25, 2018 7:19 pm
Yeah, that single tax bracket is a killer. That's why we are converting into the 22% bracket. Survivor will definitely be in the 22% single bracket, if not the 24%. I'm not sure people realize the impact of being single on their tax bracket after being MFJ for a while. It's something to consider ...
Agree, and it's actually worse with the new 2018 brackets. They did away with the "marriage penalty" so MFJ brackets are 2x the single brackets at least up to $400K.

trueblueky
Posts: 1345
Joined: Tue May 27, 2014 3:50 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by trueblueky » Wed Jul 25, 2018 8:04 pm

One Ping wrote:
Wed Jul 25, 2018 7:19 pm
GAAP wrote:
Wed Jul 25, 2018 5:46 pm
BigJohn wrote:
Wed Jul 25, 2018 5:32 pm
It doesn’t seem too difficult to get your retirement tax bracket above that once RMDs start.
Especially after the first spouse dies and the second is forced into the single tax brackets...
Yeah, that single tax bracket is a killer. That's why we are converting into the 22% bracket. Survivor will definitely be in the 22% single bracket, if not the 24%. I'm not sure people realize the impact of being single on their tax bracket after being MFJ for a while. It's something to consider ...
+1

We are in 12% bracket, but we were never at 12% marginal while working, and we will never again be at 12% after RMD, SS, or one of us die, whatever comes first. We are converting to the top of 12% this year and converted to top of 15% last year. I see nothing in this discussion to change that.

FinancialDave
Posts: 1579
Joined: Thu May 26, 2011 9:36 pm

Re: Vanguard Paper on Roth Conversion Analysis

Post by FinancialDave » Wed Jul 25, 2018 8:06 pm

FiveK wrote:
Wed Jul 25, 2018 7:13 pm
FinancialDave wrote:
Wed Jul 25, 2018 4:38 pm
Let me state it a different way. When all is said and done the optimal Roth / tIRA choice....
...
Your marginal rate tells you if you need any Roth at all, your effective tax from these accounts tells you how much.
...
The current size of your Roth / traditional funds is technically what determines whether more money should be going towards the Roth.
Before continuing point by point discussion, it may be that we have one "side" saying the sky is blue while the counterargument is that the grass is green.

From a quick read I can agree with the items above, from the perspective that (I think) you are looking at the split between traditional and Roth over an entire career. On the other hand, the wiki and Kitces articles, etc., are looking at how one might choose for a specific year's contribution.

Does that seem correct to you?
Here is a line from the Wiki that bothers me:

"The reason to use marginal tax rates in this decision is that you can make the decision separately for every dollar you invest."

First off who is going to do that. I think what the author of that WikI fails to realize is just the point I am stressing above. You would have to compare every Roth dollar to every IRA dollar spent in retirement. These IRA dollars can be spent in retirement across all tax brackets, so to know how many IRA dollars you need compared to Roth dollars means you need to know where the dollars are to be spent.

It makes no sense to me to blindly put money into a Roth just because your tax rate while working is at a specific level. I do agree with what might be the underlying premiss that the lower your working tax rate is the less you can go wrong by favoring the Roth, but if you just believe the Wiki and you thought your tax rate would be higher in retirement you would put ALL your retirement money in the Roth - absolutely the wrong thing to do for the reasons I mention above.

P.S. This is a complex subject, when you consider all the unknowns, so the authors are not totally wrong in their observations. I just don't agree with all of what they are saying. I have to admit I am a little jaded, because I have seen time and time again authors articles in books and professional journals that can't seem to even track the money from when it was earned to when it was spent. They also don't seem to realize that money is spent out of these accounts in a "number" of tax brackets and not just a "marginal one" -- it is more like marginal ONES. Now I'm ranting, so let's not go there!

P.S.S. I have written a number of articles myself in earlier years comparing marginal rates to marginal rates in and out, because this is step one to understanding the baseline and knowing what side of it you are on. I didn't realize the extent that the public is overfunding their Roths until I started writing a few articles on it, on another site I again won't mention. Like I said the marginal / marginal comparison tells you which account to favor, BUT only if you compare all dollars spent in EACH marginal bracket to where the money comes from. ALSO, tax brackets don't stay put, they go up with inflation as does the standard deduction. This may be part of the reason many people think they will be in a higher tax bracket, despite all the evidence to the contrary over the last 20 years.

Dave
I love simulated data. It turns the impossible into the possible!

Post Reply