### Author Topic: When to switch to Roth: check my math  (Read 2782 times)

#### terran

• Magnum Stache
• Posts: 3475
##### When to switch to Roth: check my math
« on: July 30, 2017, 09:29:36 PM »
I'm a firm believer in the idea that tax deferred investments are the better choice for anyone who expects to be in the same or lower tax bracket in retirement as compared to during their working years. However, my wife and I are in a low bracket (15%) and are able to contribute large amounts to tax deferred account both in terms of our ability to save and the accounts we have access to (solo 401k, 403b, 457, t.IRAs, sometimes an HSA), so it seems conceivable that we could over contribute such that we end up in a higher bracket  (25%) in retirement, so I'm trying to figure out at what point that would be.

Assumptions (not our actual situation, but the idea is the same)
>Current tax deferred balance: \$200k
>Average real rate of return (after inflation): 5%
>Current age: 30
>Age at retirement: 45
>Age at last withdrawal (death): 95

For the purposes on these calculations I will also assume that the current tax brackets remain the same when adjusted for inflation, which will allow us to make all calculations based on today's dollars. This means \$20,800 of income will be taxed at 0% (standard deduction plus two personal exemptions), the 10% bracket tops out at \$39,450 (\$20,800 + \$18,650 taxable income), and the 15% bracket tops out at \$96,700 (\$20,800 + \$75,900 taxable income).

This means to make sure I fill all of the the lower bracket(s) in retirement I want to have withdrawals of \$39,450/year, and if I don't want to go into a higher bracket in retirement I want to have withdrawals below \$96,700/year. Remember that I'm using a real rate of return, so while these numbers will actually be higher, if the tax code stays the same, this is what they'll be in inflation adjusted terms.

So using the assumed 5% real return, 60 years of withdrawals (death at 95, retire at 45), a minimum annual payment of \$39,450 and maximum annual payment of \$96,700, the present value formula tells me I want a balance at retirement between \$720,196.26 and \$1,765,347.99 in tax deferred accounts.

Now, based on those same assumptions, the resulting min/max balances, 15 years of savings (current age of 30, retire at 45), and current tax deferred balances of \$200k, the payment formula tells me I need to save between \$14,107.08/year and \$62,541.81/year to tax deferred accounts.

Things that would change these results:
>Rate of return: a higher return would make the required savings lower
>Age at retirement and current age: a higher retirement age or lower current age would make the required savings lower (more time to save and/or less time to spend)
>Age at last withdrawal (death): A lower age of death would mean lower required savings (less time to spend), although it doesn't make as big a difference as changing the current or retirement age.
>State taxes: the possibility of moving to a no income tax state would encourage trying to max out the 15% bracket in retirement as staying in the 10% would only save 5% (likely to be made up for by the state tax savings while working), but it would not encourage going into the 25% bracket in retirement since that would cost and additional 10% (unlikely to be made up for by the state tax savings while working).
> Social security or other income: the possibility of other income at some point in retirement would suggest that it might be best to leave some room in the 15% bracket for that income. I view this as icing on the cake and if I have the "misfortune" to get some extra income that pushes me into the 25% bracket, so be it.
> Current tax credits: before now we've been favoring tax deferred because it has kept us within income limits for the savers tax credit resulting on a larger marginal tax rate savings by prioritizing tax deferred. Soon we won't be able to stay under these limits no matter what we do.
> Subsidies: If you think you might get income tested subsidies (like the current healthcare subsidies) then you might want to try to keep your retirement income lower
> Significant capital gains: if you expect to have taxable investments with significant capital gains you might want to leave some room in the 15% bracket so you can harvest those gains tax free.
> Death of a spouse: This would lower the tax brackets of the surviving spouse, so this would be a reason to try to keep tax deferred balances lower.

So what do you think? Are my assumptions and calculations sound? Would you forgo tax deferred investments (by switching to Roth) if it looked like you were going to push yourself into a higher bracket in retirement, or do you think all of this has too many unknowns, so better to just keep maxing out tax deferred?

#### kenaces

• Stubble
• Posts: 248
##### Re: When to switch to Roth: check my math
« Reply #1 on: July 30, 2017, 09:53:41 PM »
So what do you think? Are my assumptions and calculations sound? Would you forgo tax deferred investments (by switching to Roth) if it looked like you were going to push yourself into a higher bracket in retirement, or do you think all of this has too many unknowns, so better to just keep maxing out tax deferred?

I don't know the answer but I am looking forward to hearing others answers. I have heard that in planning situations where decisions are close it can be helpful to have to "tax diversification" so you have more control of when and what kind of income you take.

I know there a lot of assumption that can change things but I am curious if there are some rules of thumb that can be helpful?

#### MDM

• Senior Mustachian
• Posts: 10658
##### Re: When to switch to Roth: check my math
« Reply #2 on: July 31, 2017, 01:16:53 AM »
I'm a firm believer in the idea that tax deferred investments are the better choice for anyone who expects to be in the same or lower tax bracket in retirement as compared to during their working years.
Lower, yes.  But Roth may be better for the same bracket - see Maxing out your retirement accounts for details.

Quote
So what do you think? Are my assumptions and calculations sound?
Looks good at a quick glance - very similar to what is in Investment Order for estimating withdrawal tax rates.

Quote
Would you forgo tax deferred investments (by switching to Roth) if it looked like you were going to push yourself into a higher bracket in retirement, or do you think all of this has too many unknowns, so better to just keep maxing out tax deferred?
Great question.

I think the answer depends on how one defines "if it looked like".  Because having "too much" in traditional accounts is less a problem than "too much" in Roth accounts, I'd hedge by using "conservative" projections.  Of course, even the definition of "conservative" may vary....

As noted above going to Roth seems clearly indicated if one expects to pay a higher rate in retirement.  The closer question occurs when one expects to pay the same or even slightly less.

#### Asalbeag

• Posts: 20
##### Re: When to switch to Roth: check my math
« Reply #3 on: July 31, 2017, 04:58:28 AM »
Not an answer to your original question but since you're in the 15% tax bracket I would definitely look into capital gains harvesting in your taxable accounts, up to the 25% cut off you are still in the 0% capital gains bracket. Unlike capital loss harvesting there are no wash sale rules for gains (that I'm aware of).

#### terran

• Magnum Stache
• Posts: 3475
##### Re: When to switch to Roth: check my math
« Reply #4 on: July 31, 2017, 10:00:58 AM »
I'm a firm believer in the idea that tax deferred investments are the better choice for anyone who expects to be in the same or lower tax bracket in retirement as compared to during their working years.
Lower, yes.  But Roth may be better for the same bracket - see Maxing out your retirement accounts for details.

Good point. Probably not an issue for us as I'm pretty sure my wife's 403b will allow a mega backdoor roth, so we should have more space than we can afford to fill, but for people without a ridiculous amount of tax advantaged space the fact that money in a roth is "worth" more (won't be taxed on the way out) and "costs" more (doesn't reduce your tax bill) means you can effectively put more money in a roth as compared to tax deferred.

Quote from: MDM
Quote
So what do you think? Are my assumptions and calculations sound?
Looks good at a quick glance - very similar to what is in Investment Order for estimating withdrawal tax rates.

You're right, it is similar to the investment order advice. If you're in the 10% bracket go Roth, if you're in the 25% bracket go traditional, if you're in the 15% bracket flip a coin. I guess I'm just trying to do a little better than flip a coin in the 15% even if the "calculations" (can they really be called that when the assumptions are so up in the air?) may be flawed. The good news, of course, is that it's probably not the end of the world either way.

Quote from: MDM
Quote
Would you forgo tax deferred investments (by switching to Roth) if it looked like you were going to push yourself into a higher bracket in retirement, or do you think all of this has too many unknowns, so better to just keep maxing out tax deferred?
Great question.

I think the answer depends on how one defines "if it looked like".  Because having "too much" in traditional accounts is less a problem than "too much" in Roth accounts, I'd hedge by using "conservative" projections.  Of course, even the definition of "conservative" may vary....

As noted above going to Roth seems clearly indicated if one expects to pay a higher rate in retirement.  The closer question occurs when one expects to pay the same or even slightly less.

You're so right that the definition of conservative is hard to pin down. Given that current age and tax deferred balance are fixed for any given situation, I think the biggest things try to get right are rate of return and retirement age since changing the age of last withdrawal doesn't make all that much difference in the calculations.

I think where I'm leaning is targeting somewhere around the top of the 15% in retirement (contribute \$62,541.81/year until retirement given my assumptions above) since, as you say, it's better to have "too much" tax deferred than not enough. That way the "worst" case is that we have to take so much out of tax deferred that we end up in the 25% bracket which also means we have way more money that I can see us really needing, so paying a little extra in taxes isn't going to really make any difference at that point.

The other "bad" outcome would be that returns are lower than assumed and it turns out we could have put more traditional, but since that will just means we pay the same 15% taxes then as we would now and really we'll probably have some other income to fill that up, it probably doesn't much matter either.

This does mean that we should switch some of our contributions to Roth, but it actually looks like we might have some headroom before we can no longer take advantage of the savers tax credit (pretty crazy how much you can make and still be eligible with enough tax deferred contributions), so I'll have to mess around with some numbers and see if it's worth "over contributing" to tax deferred in order to get the \$400 level of the saver's tax credit.

Not an answer to your original question but since you're in the 15% tax bracket I would definitely look into capital gains harvesting in your taxable accounts, up to the 25% cut off you are still in the 0% capital gains bracket. Unlike capital loss harvesting there are no wash sale rules for gains (that I'm aware of).

Good point. Even after we can no longer qualify for the savers tax credit (AGI \$62,000) there's still quite a bit of headroom in the 15% bracket (AGI \$96,700) so that would be a great time to do some tax gain harvesting. For our particular situation, given the amount of tax advantaged space we have we really don't/won't have that much in taxable until/unless our income goes up to a point where we're not in the 15% bracket anyway.

#### terran

• Magnum Stache
• Posts: 3475
##### Re: When to switch to Roth: check my math
« Reply #5 on: July 31, 2017, 10:26:01 AM »
I'm trying to think through when it makes sense to "break the rule" I've established above by contributing more than indicated to tax deferred in order to maintain eligibility for a tax credit (like the savers tax credit). Would it make sense to divide the potential credit (\$400) by the marginal tax  this action would push me into in retirement (say 25% federal, 0% state), minus current marginal tax (say 15% federal, 5% state), and this would be the amount over the limit set above I should be willing to go to keep the credit?

So, for this example that would be \$400/((25%+0%) - (15%+5%)) =  \$8000. So I should be willing to contribute an extra \$8000 to tax deferred for a total of \$70,541.81 if that will keep me eligible for the savers tax credit, and if I need to contribute any more than \$8000 extra to still be eligible I should revert to contributing at most \$62,541.81 to tax deferred.

Does that make sense, or has my brain gone a little fuzzy on that one?

#### secondcor521

• Magnum Stache
• Posts: 3837
• Age: 52
• Location: Boise, Idaho
• Big cattle, no hat.
##### Re: When to switch to Roth: check my math
« Reply #6 on: July 31, 2017, 11:26:52 AM »
I applaud your effort to analyze and plan long term on taxes.  It's by far many people's largest expense and bears optimizing for that reason.

1.  95 - 45 is 50, not 60.  It's probably just a typo on one of those numbers in your post, and your spreadsheet you're using is probably correct because you're probably calculating the values off of your inputs.  But if not, you should correct that error.

2.  Tax policy changes over time, sometimes in little ways, sometimes in big ways.  I'm 48 now, and off the top of my head, the following have been tax policy changes that affected me:  the ACA, the child tax credit, education credits, 529's, ESAs.  I also had an inheritance, an unplanned change in my filing status, and financial aid considerations affect my taxes now.  So I think it is good to have a plan and hedge your bets appropriately but understand that you'll need to revisit it pretty much every year.

Personally, I simplify everything to one basic rule:  I try to pay tax on anything that has a federal rate under 20%, and try to avoid paying taxes on anything that has a federal rate over 20%.  So I do my pro forma taxes in the fall, and then decide whether to contribute to a Roth, whether to contribute to a traditional IRA, whether to capital gain or capital loss harvest, and how much of a Roth conversion to do.  I use a tax program (TaxAct) so it will do all of the math and phaseouts correctly and quickly.  In the range of income I am looking at, the tax paid is augmented by various credits and deductions or losses thereof, so I just change the variables in \$1K increments and then look at the resulting increase/decrease in taxes paid and do a little Excel chart to find the sweet spots.  I keep an eye on the future, but because of item 2 above, tactically I just optimize each year.

#### MDM

• Senior Mustachian
• Posts: 10658
##### Re: When to switch to Roth: check my math
« Reply #7 on: July 31, 2017, 12:50:21 PM »
I'm trying to think through when it makes sense to "break the rule" I've established above by contributing more than indicated to tax deferred in order to maintain eligibility for a tax credit (like the savers tax credit). Would it make sense to divide the potential credit (\$400) by the marginal tax  this action would push me into in retirement (say 25% federal, 0% state), minus current marginal tax (say 15% federal, 5% state), and this would be the amount over the limit set above I should be willing to go to keep the credit?

So, for this example that would be \$400/((25%+0%) - (15%+5%)) =  \$8000. So I should be willing to contribute an extra \$8000 to tax deferred for a total of \$70,541.81 if that will keep me eligible for the savers tax credit, and if I need to contribute any more than \$8000 extra to still be eligible I should revert to contributing at most \$62,541.81 to tax deferred.

Does that make sense, or has my brain gone a little fuzzy on that one?
Capturing more of any credit (saver's, earned income, child tax, etc.) increases the marginal tax saving rate of contributions.  That can make traditional contributions better than Roth even for those in the 10% or even 0% bracket.

The comparison remains between marginal contribution vs. withdrawal rates.

Note that marginal tax rate, in this situation, does not mean "rate on the very last dollar".  It means "average rate for this amount of traditional contribution."

E.g., if one needs a \$1000 contribution to get a \$200 saver's credit, and all \$1000 reduces income in the 15% bracket, the marginal tax saving rate is (\$200 + \$150) / \$1000 = 35%.

#### terran

• Magnum Stache
• Posts: 3475
##### Re: When to switch to Roth: check my math
« Reply #8 on: July 31, 2017, 12:51:22 PM »
I applaud your effort to analyze and plan long term on taxes.  It's by far many people's largest expense and bears optimizing for that reason.

1.  95 - 45 is 50, not 60.  It's probably just a typo on one of those numbers in your post, and your spreadsheet you're using is probably correct because you're probably calculating the values off of your inputs.  But if not, you should correct that error.

Oops, right. Yes, my spreadsheet does the math for me, so the numbers should be right.

Quote from: secondcor521
2.  Tax policy changes over time, sometimes in little ways, sometimes in big ways.  I'm 48 now, and off the top of my head, the following have been tax policy changes that affected me:  the ACA, the child tax credit, education credits, 529's, ESAs.  I also had an inheritance, an unplanned change in my filing status, and financial aid considerations affect my taxes now.  So I think it is good to have a plan and hedge your bets appropriately but understand that you'll need to revisit it pretty much every year.

Personally, I simplify everything to one basic rule:  I try to pay tax on anything that has a federal rate under 20%, and try to avoid paying taxes on anything that has a federal rate over 20%.  So I do my pro forma taxes in the fall, and then decide whether to contribute to a Roth, whether to contribute to a traditional IRA, whether to capital gain or capital loss harvest, and how much of a Roth conversion to do.  I use a tax program (TaxAct) so it will do all of the math and phaseouts correctly and quickly.  In the range of income I am looking at, the tax paid is augmented by various credits and deductions or losses thereof, so I just change the variables in \$1K increments and then look at the resulting increase/decrease in taxes paid and do a little Excel chart to find the sweet spots.  I keep an eye on the future, but because of item 2 above, tactically I just optimize each year.

Good point! Other than rate of return, changing tax code is probably the biggest thing that could make all of this a pointless exercise. All we can do is go based on the information we have now.

Why have you chosen 20% as your cut off for the federal marginal rate you try to stay under?

#### terran

• Magnum Stache
• Posts: 3475
##### Re: When to switch to Roth: check my math
« Reply #9 on: July 31, 2017, 12:57:43 PM »
I'm trying to think through when it makes sense to "break the rule" I've established above by contributing more than indicated to tax deferred in order to maintain eligibility for a tax credit (like the savers tax credit). Would it make sense to divide the potential credit (\$400) by the marginal tax  this action would push me into in retirement (say 25% federal, 0% state), minus current marginal tax (say 15% federal, 5% state), and this would be the amount over the limit set above I should be willing to go to keep the credit?

So, for this example that would be \$400/((25%+0%) - (15%+5%)) =  \$8000. So I should be willing to contribute an extra \$8000 to tax deferred for a total of \$70,541.81 if that will keep me eligible for the savers tax credit, and if I need to contribute any more than \$8000 extra to still be eligible I should revert to contributing at most \$62,541.81 to tax deferred.

Does that make sense, or has my brain gone a little fuzzy on that one?
Capturing more of any credit (saver's, earned income, child tax, etc.) increases the marginal tax saving rate of contributions.  That can make traditional contributions better than Roth even for those in the 10% or even 0% bracket.

The comparison remains between marginal contribution vs. withdrawal rates.

Note that marginal tax rate, in this situation, does not mean "rate on the very last dollar".  It means "average rate for this amount of traditional contribution."

E.g., if one needs a \$1000 contribution to get a \$200 saver's credit, and all \$1000 reduces income in the 15% bracket, the marginal tax saving rate is (\$200 + \$150) / \$1000 = 35%.

Makes sense. I think this is what I was trying to get at (from another direction). So in my example, if \$8000 of additional deferral will get me the credit then the credit saves me \$400/\$8000 = 5%, plus I'll save 15% federal, and 5% state resulting in 25% overall savings which would be break even with a 25% federal 0% state marginal tax rate in retirement. So if I can get the credit with anything less than \$8000 in additional deferral then I come out ahead. Does that match up with your thinking above?

#### MDM

• Senior Mustachian
• Posts: 10658
##### Re: When to switch to Roth: check my math
« Reply #10 on: July 31, 2017, 01:03:59 PM »
Makes sense. I think this is what I was trying to get at (from another direction). So in my example, if \$8000 of additional deferral will get me the credit then the credit saves me \$400/\$8000 = 5%, plus I'll save 15% federal, and 5% state resulting in 25% overall savings which would be break even with a 25% federal 0% state marginal tax rate in retirement. So if I can get the credit with anything less than \$8000 in additional deferral then I come out ahead. Does that match up with your thinking above?
Yes, exactly.

That's the intended use of the "Cumulative" curve in the case study spreadsheet chart (over cell I75 on the 'Calculations' tab).

#### secondcor521

• Magnum Stache
• Posts: 3837
• Age: 52
• Location: Boise, Idaho
• Big cattle, no hat.
##### Re: When to switch to Roth: check my math
« Reply #11 on: July 31, 2017, 01:16:48 PM »
Good point! Other than rate of return, changing tax code is probably the biggest thing that could make all of this a pointless exercise. All we can do is go based on the information we have now.

Why have you chosen 20% as your cut off for the federal marginal rate you try to stay under?

I think long term planning is actually a great exercise and not pointless, because it gets you to think strategically about what things you can actively do to minimize your tax burden.  Just expect that you'll have to revisit your long term plan every year :-)

As for why 20%:  That is really just a very rough spitball of what I am willing to pay based on my subjective opinion of several factors:

1.  The potential for tax rates to change, up or down, in the future, by various actions of government.
2.  The potential for me to avoid, or not avoid, taxation in the future, by various actions of mine.
3.  A rough estimate of what marginal rate I expect to be over the next 20 years is probably 25-30%.
4.  The potential for me to be affected by what is sometimes called the tax torpedo - the increase in marginal rates due to losing my deductions, Social Security income, and RMDs.

Plus it's a nice round number.

I will add that I am considering going even lower than the 20% for the current tax year since it will serve as the base year for college financial aid for one of my sons.  Just as considering the effect of ACA credits, Saver's tax credits, and other credits and deductions can make the effective marginal rate on an additional dollar of taxable income, the same holds true for financial aid, which adds on top of all of that.  So between federal taxes, state taxes, reduced federal tax credits, and reduced financial aid, I could effectively be looking at "very high" (40-50%) marginal costs at relatively low income (100-200% FPL).

One of these days - and it may be this fall - I plan to pay a trusted CPA to review my numbers and give me feedback to make sure I'm thinking about all of this correctly and to see if I'm missing anything.  I expect the few hundred I pay to him will pay off either in peace of mind that my plan is pretty good and/or additional tax saving tricks I haven't thought of or discovered on my own yet.