Asymmetric risk in the Roth vs Traditional question

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Topic Author
Northern Flicker
Posts: 5780
Joined: Fri Apr 10, 2015 12:29 am

Asymmetric risk in the Roth vs Traditional question

Post by Northern Flicker » Wed Jun 24, 2020 4:44 pm

The decision of whether to make Roth or traditional 401K contributions comes up from time to time, and the discussion often does not reach a consensus, in part because the analysis is surprisingly complex.

There is however an interesting asymmetry in the consequences of making the wrong decision. Based on one's tax and income situation in retirement, there will be some traditional account balance (aggregated across all trad accounts) above which all further assets should be in Roth space. It is not easy to determine that threshold, and even if you have a good estimate for it, targeting the number accurately would require accurate estimates of future investment returns.

But it is interesting to look at the two suboptimal outcomes through the lenses of having savings that may be stretched or insufficient in retirement as one scenario, and saving more than you need as another scenario.

If you end up saving more than you need, then not having enough traditional balance just means that, despite paying more taxes than necessary on Roth contributions during your working years, you saved enough and are doing fine. Having too high of a traditional balance means that your after tax retirement assets are less than they would have been had you done more Roth contributions, which just means that you could have overfunded your retirement by a bit more than it is by having done more Roth contributions.

On the other hand, if the level of retirement savings underfunds one's retirement or has little or no margin of error, then overfunding the Roth space increases the risk of running out of funds and decreases after tax retirement assets. I don't think it is possible to have too much trad space if assets underfund a comfortable spending level.

While your target levels may be consistent with Roth contributions, there are no guarantees of length of working career or investment returns. The conclusion is the Roth contributions that could have instead been traditional contributions increase risk of a retirement shortfall for all but very high earners.
Last edited by Northern Flicker on Thu Jun 25, 2020 12:03 am, edited 2 times in total.
Risk is not a guarantor of return.

02nz
Posts: 4640
Joined: Wed Feb 21, 2018 3:17 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by 02nz » Wed Jun 24, 2020 4:47 pm

This is exactly right, and the same point I just made in another thread: "I often find that people's projections/assumptions are wildly skewed in favor of Roth ("our deficits are huge!"). They forget that if you choose Roth and turn out to be wrong, it's a costly mistake that can put a dent in your standard of living. If you choose traditional and it turns out to be wrong, it's likely because returns and/or income are higher than expected, you pay more tax than you could have but a nice problem to have!"

I think it would be good to reflect this in the wiki on Roth vs traditional. This particular issue has been discussed before, but I don't think it gets nearly the attention it deserves, it's potentially much higher-impact than "should I switch from VTSAX to VTI for the 1 bp lower ER?" or even domestic vs international. I'm sure most newbies asking the Roth vs traditional question have never thought about it; I certainly had not until I read through a lot more Roth vs traditional threads.

retired@50
Posts: 2986
Joined: Tue Oct 01, 2019 2:36 pm
Location: Living in the U.S.A.

Re: Asymmetric risk in the Roth vs Traditional question

Post by retired@50 » Wed Jun 24, 2020 5:35 pm

02nz wrote:
Wed Jun 24, 2020 4:47 pm
I'm sure most newbies asking the Roth vs traditional question have never thought about it; I certainly had not until I read through a lot more Roth vs traditional threads.
The mistake I most often see from young retirement savers is that they feel as though they will be paying higher taxes when they are 35 or 45 years old, so they choose Roth at 25 (even though they are already in a high tax bracket like 24% or higher). However, it's the tax rate during retirement that is relevant for comparison, not the tax rate during middle age.

Edit: red passage.

Regards,
Last edited by retired@50 on Sat Jun 27, 2020 8:53 am, edited 1 time in total.
This is one person's opinion. Nothing more.

02nz
Posts: 4640
Joined: Wed Feb 21, 2018 3:17 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by 02nz » Wed Jun 24, 2020 5:43 pm

retired@50 wrote:
Wed Jun 24, 2020 5:35 pm
02nz wrote:
Wed Jun 24, 2020 4:47 pm
I'm sure most newbies asking the Roth vs traditional question have never thought about it; I certainly had not until I read through a lot more Roth vs traditional threads.
The mistake I most often see from young retirement savers is that they feel as though they will be paying higher taxes when they are 35 or 45 years old, so they choose Roth at 25. However, it's the tax rate during retirement that is relevant for comparison, not the tax rate during middle age.

Regards,
To be fair, there's some justification in the thinking you cited, it's not necessarily a mistake. If I'm in the 12% bracket now and expect to be in a higher bracket in a few years because of pay raises (medical residents are kind of an extreme example of this), even with nothing in tax-deferred balances now, I might reasonably decide to contribute to Roth 401k and then later switch to traditional once at 22% or higher, to get more tax diversification. I think for most people the best way to get that diversification is by combining trad'l 401k and Roth IRA (and the point about asymmetric risk applies here - what if those anticipated raises or career advancements don't work out, for example), but starting out with Roth when one's income is at its lowest is not an unreasonable choice, either.

User avatar
Watty
Posts: 19630
Joined: Wed Oct 10, 2007 3:55 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by Watty » Wed Jun 24, 2020 6:17 pm

Northern Flicker wrote:
Wed Jun 24, 2020 4:44 pm
But it is interesting to look at the two suboptimal outcomes through the lenses of having savings that may be stretched or insufficient in retirement as one scenario, and saving mire than you need as another scenario.
I agree with your post.

An additional factor is that if you find that your are accumulating too much in a traditional account then you have the option of;

1) Retiring a few years early in a lower tax bracket.

2) Doing Roth conversions.

If everything goes well and you end up in a high retirement tax bracket part of that may also be that you got lucky and did not run into any unexpected setbacks. Especially when I was going through my 50s I saw more people than I would have expected run into financial setbacks with careers, layoffs, health problems, the death of a spouse, divorces, kids with expensive problem, etc that caused them to not be as well off in retirement as they expected.

jhawktx
Posts: 199
Joined: Thu May 19, 2011 8:15 am

Re: Asymmetric risk in the Roth vs Traditional question

Post by jhawktx » Wed Jun 24, 2020 6:34 pm

This message board does seem to go overboard on pushing high Roth balances. Good points made by OP. Also, if you have a nice well paying job today with a rosy outlook but are 10-20 years from retirement, there are no guarantees your rosy financial outlook will pan out as hoped/projected.

dboeger1
Posts: 216
Joined: Fri Jan 13, 2017 7:32 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by dboeger1 » Wed Jun 24, 2020 6:48 pm

Watty wrote:
Wed Jun 24, 2020 6:17 pm
An additional factor is that if you find that your are accumulating too much in a traditional account then you have the option of;

1) Retiring a few years early in a lower tax bracket.

2) Doing Roth conversions.
I think this is an extremely important point that arguably applies more to the types of people trying to optimize between both account types than they seem to recognize. If you're a young, low-income worker, Roth is a no-brainer. Likewise, if you earn a high income, maybe a little older, are behind on your retirement savings and so RMDs and retirement tax brackets are not a major concern, and you aren't looking to optimize every last penny of tax-advantaged space, traditional is a no-brainer because it has guaranteed huge benefits up front with a high marginal tax rate. It's the extreme retirement savings optimizers who typically care about this stuff, and even most of them acknowledge they don't have a crystal ball and likely end up splitting the difference by using both. After all, using both is often the most natural thing because of how employee matches generally go to traditional 401(k) accounts anyway, and income limits force higher-income IRA users to use the back door Roth method. So trying to perfectly optimize the ratio of these account types is splitting hairs, and mostly only applies to the most extreme of retirement savers, such as FIRE people looking to reach their retirement goals early. Traditional retirees who save 10%-20% of their incomes for 30-40 years tend to fall into the obvious Roth, traditional, or split-the-difference categories.

So for those extreme retirement savers for whom the difference actually matters because they're on a war path to save up huge amounts in retirement accounts as quickly as possible, they need to consider not only the optimal ratio, but also ways to game the system. For example, if you earn a high income at the start of your career, you can defer the taxes in a traditional account, "retire early" (or just take a few years off), do Roth conversions up a low tax bracket, and basically end up with huge amounts of very long-term Roth savings having paid very little if any taxes on them. If one intends to start a business, they don't even need to take those years off. They could spend a couple of years building the business with minimal earnings by doing unpaid work for referrals, and then start generating profits after the conversions. In effect, the government would be giving you a retirement savings tax break to build your business from scratch, in addition to all the other business-friendly tax incentives. Many FIRE types aspire to be entrepreneurs, so it seems like an obvious strategy. That alone is likely to have a way bigger impact when it comes to tax reduction than any small adjustment to the ratio of Roth and traditional accounts. That effectively adds a ton of value to traditional retirement accounts for young people with high incomes who can see themselves retiring early, because they can exploit that flexibility by timing their Roth conversions to minimize taxes.

fctu
Posts: 43
Joined: Wed Jan 31, 2018 10:23 am

Re: Asymmetric risk in the Roth vs Traditional question

Post by fctu » Wed Jun 24, 2020 6:58 pm

We cannot assume that tax rates will be the same in retirement. Tax policy is changed regularly.

If taxes are lowered to zero in retirement, Roth will do ok but traditional will do better.

If marginal tax rate is 100% in retirement, Roth does well and traditional completely fails.

I like to be prepared for both extremes and any point in between.

retired@50
Posts: 2986
Joined: Tue Oct 01, 2019 2:36 pm
Location: Living in the U.S.A.

Re: Asymmetric risk in the Roth vs Traditional question

Post by retired@50 » Wed Jun 24, 2020 7:01 pm

02nz wrote:
Wed Jun 24, 2020 5:43 pm
retired@50 wrote:
Wed Jun 24, 2020 5:35 pm
02nz wrote:
Wed Jun 24, 2020 4:47 pm
I'm sure most newbies asking the Roth vs traditional question have never thought about it; I certainly had not until I read through a lot more Roth vs traditional threads.
The mistake I most often see from young retirement savers is that they feel as though they will be paying higher taxes when they are 35 or 45 years old, so they choose Roth at 25. However, it's the tax rate during retirement that is relevant for comparison, not the tax rate during middle age.

Regards,
To be fair, there's some justification in the thinking you cited, it's not necessarily a mistake. If I'm in the 12% bracket now and expect to be in a higher bracket in a few years because of pay raises (medical residents are kind of an extreme example of this), even with nothing in tax-deferred balances now, I might reasonably decide to contribute to Roth 401k and then later switch to traditional once at 22% or higher, to get more tax diversification. I think for most people the best way to get that diversification is by combining trad'l 401k and Roth IRA (and the point about asymmetric risk applies here - what if those anticipated raises or career advancements don't work out, for example), but starting out with Roth when one's income is at its lowest is not an unreasonable choice, either.
Agreed if the current bracket is low, but we often see young savers already in the 24% bracket and above, who have still chosen Roth. I have made an edit to my earlier remarks.

Regards,
Last edited by retired@50 on Sat Jun 27, 2020 8:54 am, edited 1 time in total.
This is one person's opinion. Nothing more.

Longdog
Posts: 1473
Joined: Sun Feb 09, 2014 6:56 pm
Location: Philadelphia

Re: Asymmetric risk in the Roth vs Traditional question

Post by Longdog » Wed Jun 24, 2020 7:22 pm

I feel like there are a lot of unstated assumptions that everyone responding seems to understand but are making it difficult for me (and possibly others) to fully appreciate the conclusions. For example, is there an assumption that the "perfect" allocation will spend your last dollar the moment before you take your last breath? Is there an assumption that the entirety of your retirement portfolio consists of either traditional pre-tax accounts or Roth accounts, but not amounts held in taxable accounts? Are there assumptions about Roth conversions? Are there assumptions about low marginal tax rate years between retirement and collecting social security or beginning RMDs?

I'd sure like to understand.
Steve

02nz
Posts: 4640
Joined: Wed Feb 21, 2018 3:17 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by 02nz » Wed Jun 24, 2020 7:28 pm

Longdog wrote:
Wed Jun 24, 2020 7:22 pm
I feel like there are a lot of unstated assumptions that everyone responding seems to understand but are making it difficult for me (and possibly others) to fully appreciate the conclusions. For example, is there an assumption that the "perfect" allocation will spend your last dollar the moment before you take your last breath? Is there an assumption that the entirety of your retirement portfolio consists of either traditional pre-tax accounts or Roth accounts, but not amounts held in taxable accounts? Are there assumptions about Roth conversions? Are there assumptions about low marginal tax rate years between retirement and collecting social security or beginning RMDs?

I'd sure like to understand.
I don't think any of that is "assumed." The point made in this thread isn't that traditional is better than Roth or vice versa for some given set of assumptions; it's that you need to think about the downside of any Roth vs. traditional bet, the "what if I'm wrong?" The downsides are asymmetrical (betting Roth and being wrong means a potentially much worse outcome than betting traditional and being wrong), and that argues for tilting toward traditional when in doubt or when it's a wash (e.g., expectations about one's personal tax rates now vs in retirement are roughly the same). That doesn't necessarily mean 100% traditional, just tilting, hence the frequent recommendation here to max both a traditional 401k and Roth IRA, which (assuming same investments) will mean roughly a 3:1 traditional to Roth ratio. (That's not to suggest there's one "right" ratio, either.)

There is one important assumption, though: that while tax rates can change, they will remain relatively stable in the big scheme of things. There are a lot of people who say "but rates must go up, we have so much debt." Without speculating about future changes (which is against forum policy), I'd just point out that rates haven't changed that often, and the changes are relatively small. For example, the top federal income tax rate has remained between 35 and 39.6% for almost 3 decades. As a poster wrote in another thread: "[O]ne does not need to know precisely what future tax rates will be. In many instances, huge changes to tax rates would be necessary to shift the 'optimal' strategy from one type of account to another. For instance, a household in the 22% bracket that expects to be in the 12% bracket in retirement should clearly prefer tax-deferred accounts. Even if the 12% tax rate increases, as long as it remains lower than 22%, the household will still be better off with tax-deferred than Roth."

Sorry if that doesn't really answer your question, but personal finance is personal. There are many considerations that are unique to every individual, and these considerations interact in complex ways, which is why we usually ask for posters to give us as much information as possible about their circumstances.

User avatar
TomatoTomahto
Posts: 10560
Joined: Mon Apr 11, 2011 1:48 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by TomatoTomahto » Wed Jun 24, 2020 8:19 pm

Longdog wrote:
Wed Jun 24, 2020 7:22 pm
I feel like there are a lot of unstated assumptions that everyone responding seems to understand but are making it difficult for me (and possibly others) to fully appreciate the conclusions.
For example, is there an assumption that the "perfect" allocation will spend your last dollar the moment before you take your last breath? Goodness gracious, I sure hope not.

Is there an assumption that the entirety of your retirement portfolio consists of either traditional pre-tax accounts or Roth accounts, but not amounts held in taxable accounts? No, I think many think the tax bomb is detonated by a combination of SS, pensions, RMDs, the potential of filing single, and the taxable accounts throwing off roughly 2% annually.


Are there assumptions about Roth conversions?

Are there assumptions about low marginal tax rate years between retirement and collecting social security or beginning RMDs? I wish, but I guess it’s a “first world problem.”

I'd sure like to understand.
Okay, I get it; I won't be political or controversial. The Earth is flat.

User avatar
Ben Mathew
Posts: 1041
Joined: Tue Mar 13, 2018 11:41 am
Location: Seattle
Contact:

Re: Asymmetric risk in the Roth vs Traditional question

Post by Ben Mathew » Wed Jun 24, 2020 8:23 pm

A valid point. The progressive structure of the income tax code means that if things turned out badly (poor returns, layoffs), you're likely to be in a lower income tax bracket at retirement. So you get a break. That's a strong argument for traditional for people on the margin between traditional and Roth, especially if one is confident that the income tax regime won't change between now and retirement. But if one is worried that the income tax regime will shift, Roth offers the ability to remove the tax risk. You prepay your taxes today at known rates, and you don't have to worry about what your marginal tax rate will be decades from now in retirement.

Topic Author
Northern Flicker
Posts: 5780
Joined: Fri Apr 10, 2015 12:29 am

Re: Asymmetric risk in the Roth vs Traditional question

Post by Northern Flicker » Thu Jun 25, 2020 12:20 am

Longdog wrote:
Wed Jun 24, 2020 7:22 pm
I feel like there are a lot of unstated assumptions that everyone responding seems to understand but are making it difficult for me (and possibly others) to fully appreciate the conclusions. For example, is there an assumption that the "perfect" allocation will spend your last dollar the moment before you take your last breath? Is there an assumption that the entirety of your retirement portfolio consists of either traditional pre-tax accounts or Roth accounts, but not amounts held in taxable accounts? Are there assumptions about Roth conversions? Are there assumptions about low marginal tax rate years between retirement and collecting social security or beginning RMDs?
Those questions need to be answered if you want to try to find the optimal answer, and are part of why answering the question optimally is complex. But my concern in the thread is with risk, not expected outcome. However you answer the above and other questions are fine if you want to try to optimize the expected case. But a different way to look at it is from the risk that assumptions you make fail in ways that have negative consequences.

Savers who get closer to retirement in good shape have less risk of failure, and may choose to optimize for expected outcome instead of minimizing risk.
Risk is not a guarantor of return.

User avatar
FiveK
Posts: 9226
Joined: Sun Mar 16, 2014 2:43 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by FiveK » Thu Jun 25, 2020 1:13 am

02nz wrote:
Wed Jun 24, 2020 4:47 pm
This is exactly right, and the same point I just made in another thread: "I often find that people's projections/assumptions are wildly skewed in favor of Roth ("our deficits are huge!"). They forget that if you choose Roth and turn out to be wrong, it's a costly mistake that can put a dent in your standard of living. If you choose traditional and it turns out to be wrong, it's likely because returns and/or income are higher than expected, you pay more tax than you could have but a nice problem to have!"

I think it would be good to reflect this in the wiki on Roth vs traditional. This particular issue has been discussed before, but I don't think it gets nearly the attention it deserves, it's potentially much higher-impact than "should I switch from VTSAX to VTI for the 1 bp lower ER?" or even domestic vs international. I'm sure most newbies asking the Roth vs traditional question have never thought about it; I certainly had not until I read through a lot more Roth vs traditional threads.
In the Estimating future marginal tax rate section of the t vs. R wiki, there is "If you pick traditional and that ends up being wrong it will be because you have "too much money" - not the worst problem. If you pick Roth and that ends up being wrong it will be because you have "too little money" - that can be a problem. Thus using traditional is a "safer" choice."

There is also a lot of other information in that wiki, so it's possible to overlook that one sentence, but at least it's in there.

02nz
Posts: 4640
Joined: Wed Feb 21, 2018 3:17 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by 02nz » Thu Jun 25, 2020 9:51 am

FiveK wrote:
Thu Jun 25, 2020 1:13 am
02nz wrote:
Wed Jun 24, 2020 4:47 pm
This is exactly right, and the same point I just made in another thread: "I often find that people's projections/assumptions are wildly skewed in favor of Roth ("our deficits are huge!"). They forget that if you choose Roth and turn out to be wrong, it's a costly mistake that can put a dent in your standard of living. If you choose traditional and it turns out to be wrong, it's likely because returns and/or income are higher than expected, you pay more tax than you could have but a nice problem to have!"

I think it would be good to reflect this in the wiki on Roth vs traditional. This particular issue has been discussed before, but I don't think it gets nearly the attention it deserves, it's potentially much higher-impact than "should I switch from VTSAX to VTI for the 1 bp lower ER?" or even domestic vs international. I'm sure most newbies asking the Roth vs traditional question have never thought about it; I certainly had not until I read through a lot more Roth vs traditional threads.
In the Estimating future marginal tax rate section of the t vs. R wiki, there is "If you pick traditional and that ends up being wrong it will be because you have "too much money" - not the worst problem. If you pick Roth and that ends up being wrong it will be because you have "too little money" - that can be a problem. Thus using traditional is a "safer" choice."

There is also a lot of other information in that wiki, so it's possible to overlook that one sentence, but at least it's in there.
Good catch, I missed that despite linking to that wiki many times! :oops:

Topic Author
Northern Flicker
Posts: 5780
Joined: Fri Apr 10, 2015 12:29 am

Re: Asymmetric risk in the Roth vs Traditional question

Post by Northern Flicker » Thu Jun 25, 2020 1:47 pm

FiveK wrote:
Thu Jun 25, 2020 1:13 am
02nz wrote:
Wed Jun 24, 2020 4:47 pm
This is exactly right, and the same point I just made in another thread: "I often find that people's projections/assumptions are wildly skewed in favor of Roth ("our deficits are huge!"). They forget that if you choose Roth and turn out to be wrong, it's a costly mistake that can put a dent in your standard of living. If you choose traditional and it turns out to be wrong, it's likely because returns and/or income are higher than expected, you pay more tax than you could have but a nice problem to have!"

I think it would be good to reflect this in the wiki on Roth vs traditional. This particular issue has been discussed before, but I don't think it gets nearly the attention it deserves, it's potentially much higher-impact than "should I switch from VTSAX to VTI for the 1 bp lower ER?" or even domestic vs international. I'm sure most newbies asking the Roth vs traditional question have never thought about it; I certainly had not until I read through a lot more Roth vs traditional threads.
In the Estimating future marginal tax rate section of the t vs. R wiki, there is "If you pick traditional and that ends up being wrong it will be because you have "too much money" - not the worst problem. If you pick Roth and that ends up being wrong it will be because you have "too little money" - that can be a problem. Thus using traditional is a "safer" choice."

There is also a lot of other information in that wiki, so it's possible to overlook that one sentence, but at least it's in there.
The risk is bigger than just whether you save enough. Roth assets have a higher after-tax volatility. This implies that paying the tax up front for assets to be Roth amplifies sequence of return risk for a retiree. Said another way, with traditional assets, the government shoulders some of the sequence of returns risk.

While you can view a bad sequence of returns as just an instance of having savings fall short, it is more subtle than that. It becomes a savings deficit if sequence of returns risk materializes, but the risk is always present for a retiree. Because Roth assets amplify risk, a retiree needs to use a more conservative allocation when the percentage of Roth assets is higher to have the same level of risk as when the percentage of Roth assets is lower. This may lead to a lower expected return of the portfolio in retirement, increasing inflation risk and increasing the risk of portfolio failure.
Risk is not a guarantor of return.

theTRA
Posts: 26
Joined: Thu May 14, 2020 4:35 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by theTRA » Thu Jun 25, 2020 2:59 pm

Northern Flicker wrote:
Thu Jun 25, 2020 1:47 pm
FiveK wrote:
Thu Jun 25, 2020 1:13 am
02nz wrote:
Wed Jun 24, 2020 4:47 pm
This is exactly right, and the same point I just made in another thread: "I often find that people's projections/assumptions are wildly skewed in favor of Roth ("our deficits are huge!"). They forget that if you choose Roth and turn out to be wrong, it's a costly mistake that can put a dent in your standard of living. If you choose traditional and it turns out to be wrong, it's likely because returns and/or income are higher than expected, you pay more tax than you could have but a nice problem to have!"

I think it would be good to reflect this in the wiki on Roth vs traditional. This particular issue has been discussed before, but I don't think it gets nearly the attention it deserves, it's potentially much higher-impact than "should I switch from VTSAX to VTI for the 1 bp lower ER?" or even domestic vs international. I'm sure most newbies asking the Roth vs traditional question have never thought about it; I certainly had not until I read through a lot more Roth vs traditional threads.
In the Estimating future marginal tax rate section of the t vs. R wiki, there is "If you pick traditional and that ends up being wrong it will be because you have "too much money" - not the worst problem. If you pick Roth and that ends up being wrong it will be because you have "too little money" - that can be a problem. Thus using traditional is a "safer" choice."

There is also a lot of other information in that wiki, so it's possible to overlook that one sentence, but at least it's in there.
The risk is bigger than just whether you save enough. Roth assets have a higher after-tax volatility. This implies that paying the tax up front for assets to be Roth amplifies sequence of return risk for a retiree. Said another way, with traditional assets, the government shoulders some of the sequence of returns risk.

While you can view a bad sequence of returns as just an instance of having savings fall short, it is more subtle than that. It becomes a savings deficit if sequence of returns risk materializes, but the risk is always present for a retiree. Because Roth assets amplify risk, a retiree needs to use a more conservative allocation when the percentage of Roth assets is higher to have the same level of risk as when the percentage of Roth assets is lower. This may lead to a lower expected return of the portfolio in retirement, increasing inflation risk and increasing the risk of portfolio failure.
Good post and fascinating to see the nuances here.

redmaw
Posts: 41
Joined: Mon Apr 22, 2019 7:20 am

Re: Asymmetric risk in the Roth vs Traditional question

Post by redmaw » Fri Jun 26, 2020 7:37 am

Northern Flicker wrote:
Thu Jun 25, 2020 1:47 pm

... Roth assets have a higher after-tax volatility...
Can you explain this assertion? How are roth assets more volatile? I assume the reasoning is 100 goes up to 120, in roth thats a 20% increase and $20, in trad (assuming 22% tax rate) thats 78 aftertax dollars before the increase and 93.6 after the increase see it only changed be $15.6 instead of $20. But 15.6 is still 20% of the after tax ira value, that is the after tax value of both account changed by 20%, the raw dollar amount is different because the traditional account is worth less after tax. Am I missing something?

aristotelian
Posts: 7566
Joined: Wed Jan 11, 2017 8:05 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by aristotelian » Fri Jun 26, 2020 7:53 am

The way I see it, if it turns out that I am in a higher tax bracket in retirement, it means I "won the game." Being in a higher tax bracket is a good problem to have. The most likely scenario I see this happening is inheritance which gives me income that I am not currently counting on. In that scenario, I won't really regret having chosen traditional. I will lose more of my 401k to tax than anticipated but that is only because I have substantially more assets to draw from.

conservativeinvestor
Posts: 45
Joined: Tue Feb 05, 2019 12:14 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by conservativeinvestor » Fri Jun 26, 2020 8:52 am

This all seems very confusing.

Aren't the only things that really matter your current tax rate and future tax rate?

Lets try some math:
Salary 100,000 per year
Current tax rate 25%
Retirement tax rate 25%
20 years of contributions
7% rate of return

Traditional Contribution: 18% or 18,000 per year
Roth Contribution: 13.5% or 13,500 per year
BUT: both of these scenarios will take the same amount out of your paycheck so you will receive the same take home pay with either of these choices.

What is the better choice?
Lets do some more calculating:
Traditional ending balance before tax: 781,389
Roth ending balance: 586,042

Traditional ending balance after tax: 586,042
Roth ending balance: 586,042

If your tax bracket never changes it doesn't matter which one you choose.
So all you need to know is: What is your tax bracket now and what will it be when you retire?

jason2459
Posts: 114
Joined: Wed May 06, 2020 7:59 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by jason2459 » Fri Jun 26, 2020 10:04 am

Well, do that same thing but with someone maxing out the 401k and Roth contributions. With those assumptions.

02nz
Posts: 4640
Joined: Wed Feb 21, 2018 3:17 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by 02nz » Fri Jun 26, 2020 10:12 am

conservativeinvestor wrote:
Fri Jun 26, 2020 8:52 am
If your tax bracket never changes it doesn't matter which one you choose.
So all you need to know is: What is your tax bracket now and what will it be when you retire?
Read narrowly, that's correct (although marginal rate is what matters and can often be different from the tax bracket, especially in retirement). Your post did nothing more than prove the commutative property of multiplication. :P

But there are a ton of factors that influence the tax rate (not just what's in the tax code, but what applies to me based on my situation). And the big factor you seem to have overlooked: once you retire, you have no more earnings and you now have a $12200 (per person) standard deduction plus the 10 and 12% brackets to fill, EVERY YEAR. At current rates, a married couple with no other income can withdraw or do Roth conversions on more than $100K of their tax-deferred balance EVERY YEAR and pay an average of 9% federal income tax. The earlier one retires, the more powerful this is, but even for a couple retiring at 65, deferring SS to age 70 allows them to draw at least half a million at the lower rate, saving them $80K in taxes (vs 25%).

More to the point of this thread: Until I know I've "won the game," if it's a wash and/or there's a lot of uncertainty, the asymmetrical nature of the risks suggests I should tilt toward traditional. That doesn't necessarily mean 100%, just tilt.
Last edited by 02nz on Fri Jun 26, 2020 10:46 am, edited 1 time in total.

randomguy
Posts: 9019
Joined: Wed Sep 17, 2014 9:00 am

Re: Asymmetric risk in the Roth vs Traditional question

Post by randomguy » Fri Jun 26, 2020 10:35 am

02nz wrote:
Fri Jun 26, 2020 10:12 am
conservativeinvestor wrote:
Fri Jun 26, 2020 8:52 am
If your tax bracket never changes it doesn't matter which one you choose.
So all you need to know is: What is your tax bracket now and what will it be when you retire?
Read narrowly, that's correct (although marginal rate is what matters and can often be different from the tax bracket, especially in retirement).

But there are a ton of factors that influence the tax rate (not just what's in the tax code, but what applies to me based on my situation). And the big factor you seem to have overlooked: once you retire, you have no more earnings and you now have a $12200 (per person) standard deduction plus the 10 and 12% brackets to fill, EVERY YEAR. At current rates, a married couple with no other income can withdraw or do Roth conversions on more than $100K of their tax-deferred balance EVERY YEAR and pay an average of 9% federal income tax. The earlier one retires, the more powerful this is, but even for a couple retiring at 65, deferring SS to age 70 allows them to draw at least half a million at the lower rate, saving them $80K in taxes (vs 25%).

More to the point of this thread: Until I know I've "won the game," if it's a wash and/or there's a lot of uncertainty, the asymmetrical nature of the risks suggests I should tilt toward traditional. That doesn't necessarily mean 100%, just tilt.
Yes tax brackets don't matter. It is the effective tax (i.e. taxes paid/withdrawals) paid on both contributions and withdrawals. But good luck figuring out that number. You have to make a bunch of guesses about SS taxation, IRMAA, state taxes, tax code changes, and so on. For most people without pensions, sticking with traditional til you have 1-2 million seems to work out best. That is enough to fill up those low brackets.

02nz
Posts: 4640
Joined: Wed Feb 21, 2018 3:17 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by 02nz » Fri Jun 26, 2020 10:41 am

randomguy wrote:
Fri Jun 26, 2020 10:35 am
Yes tax brackets don't matter. It is the effective tax (i.e. taxes paid/withdrawals) paid on both contributions and withdrawals. But good luck figuring out that number. You have to make a bunch of guesses about SS taxation, IRMAA, state taxes, tax code changes, and so on. For most people without pensions, sticking with traditional til you have 1-2 million seems to work out best. That is enough to fill up those low brackets.
Be careful or you'll get into another marginal vs effective debate! :happy

You're correct that there's uncertainty and so most people just have to make some guesses. The point of this thread, though, is that there's a factor beyond SS taxes, IRMAA, etc., that many people have not thought of when they do that guessing, and so when there's a lot of uncertainty or it looks like a wash, one would do well do tilt a bit more toward traditional.

KlangFool
Posts: 16561
Joined: Sat Oct 11, 2008 12:35 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by KlangFool » Fri Jun 26, 2020 10:44 am

conservativeinvestor wrote:
Fri Jun 26, 2020 8:52 am
This all seems very confusing.

Aren't the only things that really matter your current tax rate and future tax rate?

Lets try some math:
Salary 100,000 per year
Current tax rate 25%
Retirement tax rate 25%
20 years of contributions
7% rate of return

Traditional Contribution: 18% or 18,000 per year
Roth Contribution: 13.5% or 13,500 per year
BUT: both of these scenarios will take the same amount out of your paycheck so you will receive the same take home pay with either of these choices.

What is the better choice?
Lets do some more calculating:
Traditional ending balance before tax: 781,389
Roth ending balance: 586,042

Traditional ending balance after tax: 586,042
Roth ending balance: 586,042

If your tax bracket never changes it doesn't matter which one you choose.
So all you need to know is: What is your tax bracket now and what will it be when you retire?
conservativeinvestor,

<<Current tax rate 25%
Retirement tax rate 25%>>

This is wrong!

<<Traditional ending balance before tax: 781,389
Roth ending balance: 586,042>>

4% of 781,389 = 31K per year. How does a person pay 25% taxes with 31K per year?

In order to generate 100K per year in retirement, the person would need 25X 100K = 2.5 million in the tax-deferred account.

The math is simply wrong!

KlangFool

conservativeinvestor
Posts: 45
Joined: Tue Feb 05, 2019 12:14 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by conservativeinvestor » Fri Jun 26, 2020 11:06 am

KlangFool wrote:
Fri Jun 26, 2020 10:44 am
conservativeinvestor wrote:
Fri Jun 26, 2020 8:52 am
This all seems very confusing.

Aren't the only things that really matter your current tax rate and future tax rate?

Lets try some math:
Salary 100,000 per year
Current tax rate 25%
Retirement tax rate 25%
20 years of contributions
7% rate of return

Traditional Contribution: 18% or 18,000 per year
Roth Contribution: 13.5% or 13,500 per year
BUT: both of these scenarios will take the same amount out of your paycheck so you will receive the same take home pay with either of these choices.

What is the better choice?
Lets do some more calculating:
Traditional ending balance before tax: 781,389
Roth ending balance: 586,042

Traditional ending balance after tax: 586,042
Roth ending balance: 586,042

If your tax bracket never changes it doesn't matter which one you choose.
So all you need to know is: What is your tax bracket now and what will it be when you retire?
conservativeinvestor,

<<Current tax rate 25%
Retirement tax rate 25%>>

This is wrong!

<<Traditional ending balance before tax: 781,389
Roth ending balance: 586,042>>

4% of 781,389 = 31K per year. How does a person pay 25% taxes with 31K per year?

In order to generate 100K per year in retirement, the person would need 25X 100K = 2.5 million in the tax-deferred account.

The math is simply wrong!

KlangFool
The calculations came from an online calculator. I have tried a couple different ones and always end up with similar results. if the tax rate never changes you always end up with the traditional and roth providing an equal retirement income.

So how is one to account for all of these variables and make an informed decision on which is the preferred path if no one can agree on how the calculations should be performed?

yog
Posts: 14
Joined: Wed Jan 15, 2020 12:57 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by yog » Fri Jun 26, 2020 11:09 am

While the accumulation phase uses the top of the tax bracket for the savings or deduction, the withdrawal phase starts the tax calculation from the bottom, potentially at 0%. This misunderstanding is what trips up many people in the middle brackets today, and it's easy to misuse the online calculators. As KlangFool noted, you have to have a lot of taxable income generating assets or other taxable income from pensions, trusts, etc., before the Roth makes financial sense at these brackets. It's a high hurdle, and not one easily met on an income of $100k.

Using the numbers above and assuming a 4% withdrawal, the Roth withdrawal is about $23,422, and the marginal tax rate is indeed 0%. The withdrawal from the traditional gives us about $31,256, but starting from $0 in income, the marginal tax rate is 12% and not the 25% we've already locked in with the Roth. If we calculate taxes due from a $0 base, we have $2,093 in federal tax, and the effective tax rate is actually %6.7 - much better than the 25% we locked in with the Roth. The traditional ends up giving one 24.5% more spending power annually during retirement ($29,163 vs. $23,422).

02nz
Posts: 4640
Joined: Wed Feb 21, 2018 3:17 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by 02nz » Fri Jun 26, 2020 11:17 am

conservativeinvestor wrote:
Fri Jun 26, 2020 11:06 am
KlangFool wrote:
Fri Jun 26, 2020 10:44 am
conservativeinvestor wrote:
Fri Jun 26, 2020 8:52 am
This all seems very confusing.

Aren't the only things that really matter your current tax rate and future tax rate?

Lets try some math:
Salary 100,000 per year
Current tax rate 25%
Retirement tax rate 25%
20 years of contributions
7% rate of return

Traditional Contribution: 18% or 18,000 per year
Roth Contribution: 13.5% or 13,500 per year
BUT: both of these scenarios will take the same amount out of your paycheck so you will receive the same take home pay with either of these choices.

What is the better choice?
Lets do some more calculating:
Traditional ending balance before tax: 781,389
Roth ending balance: 586,042

Traditional ending balance after tax: 586,042
Roth ending balance: 586,042

If your tax bracket never changes it doesn't matter which one you choose.
So all you need to know is: What is your tax bracket now and what will it be when you retire?
conservativeinvestor,

<<Current tax rate 25%
Retirement tax rate 25%>>

This is wrong!

<<Traditional ending balance before tax: 781,389
Roth ending balance: 586,042>>

4% of 781,389 = 31K per year. How does a person pay 25% taxes with 31K per year?

In order to generate 100K per year in retirement, the person would need 25X 100K = 2.5 million in the tax-deferred account.

The math is simply wrong!

KlangFool
The calculations came from an online calculator. I have tried a couple different ones and always end up with similar results. if the tax rate never changes you always end up with the traditional and roth providing an equal retirement income.

So how is one to account for all of these variables and make an informed decision on which is the preferred path if no one can agree on how the calculations should be performed?
There's not one "correct" calculation - the variables are too complex for that. You need to post your personal situation and then we can give recommendations. The most important factors are:

- Current marginal tax rate, state/federal
- Current size of tax-deferred portfolio
- Age
- Filing status
- Likelihood of early retirement (even if only by a few years)
- If in state with income tax, likelihood of moving to one with lower or no income tax (e.g., FL, TX)
- Pension and expected payment
Last edited by 02nz on Fri Jun 26, 2020 11:19 am, edited 1 time in total.

Walkure
Posts: 219
Joined: Tue Apr 11, 2017 9:59 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by Walkure » Fri Jun 26, 2020 11:19 am

02nz wrote:
Fri Jun 26, 2020 10:41 am
Be careful or you'll get into another marginal vs effective debate! :happy
Wait for it...
yog wrote:
Fri Jun 26, 2020 11:09 am
While the accumulation phase uses the top of the tax bracket for the savings or deduction, the withdrawal phase starts the tax calculation from the bottom, potentially at 0%. This misunderstanding is what trips up many people in the middle brackets today, and it's easy to misuse the online calculators... If we calculate taxes due from a $0 base, we have $2,093 in federal tax, and the effective tax rate is actually %6.7 - much better than the 25% we locked in with the Roth. The traditional ends up giving one 24.5% more spending power annually during retirement ($29,163 vs. $23,422).
And there it is. :oops:

https://www.bogleheads.org/wiki/Traditi ... onceptions

KlangFool
Posts: 16561
Joined: Sat Oct 11, 2008 12:35 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by KlangFool » Fri Jun 26, 2020 11:23 am

conservativeinvestor wrote:
Fri Jun 26, 2020 11:06 am


The calculations came from an online calculator. I have tried a couple different ones and always end up with similar results. if the tax rate never changes you always end up with the traditional and roth providing an equal retirement income.

So how is one to account for all of these variables and make an informed decision on which is the preferred path if no one can agree on how the calculations should be performed?
conservativeinvestor,

GIGO -> Garbage in Garbage Out.

A) What is the tax rate for 100K of income? 25%?

B) What is the tax rate for 31K of income? 10%/12%?

The tax rate will always change.

KlangFool

KlangFool
Posts: 16561
Joined: Sat Oct 11, 2008 12:35 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by KlangFool » Fri Jun 26, 2020 11:26 am

conservativeinvestor wrote:
Fri Jun 26, 2020 11:06 am
KlangFool wrote:
Fri Jun 26, 2020 10:44 am
conservativeinvestor wrote:
Fri Jun 26, 2020 8:52 am
This all seems very confusing.

Aren't the only things that really matter your current tax rate and future tax rate?

Lets try some math:
Salary 100,000 per year
Current tax rate 25%
Retirement tax rate 25%
20 years of contributions
7% rate of return

Traditional Contribution: 18% or 18,000 per year
Roth Contribution: 13.5% or 13,500 per year
BUT: both of these scenarios will take the same amount out of your paycheck so you will receive the same take home pay with either of these choices.

What is the better choice?
Lets do some more calculating:
Traditional ending balance before tax: 781,389
Roth ending balance: 586,042

Traditional ending balance after tax: 586,042
Roth ending balance: 586,042

If your tax bracket never changes it doesn't matter which one you choose.
So all you need to know is: What is your tax bracket now and what will it be when you retire?
conservativeinvestor,

<<Current tax rate 25%
Retirement tax rate 25%>>

This is wrong!

<<Traditional ending balance before tax: 781,389
Roth ending balance: 586,042>>

4% of 781,389 = 31K per year. How does a person pay 25% taxes with 31K per year?

In order to generate 100K per year in retirement, the person would need 25X 100K = 2.5 million in the tax-deferred account.

The math is simply wrong!

KlangFool
The calculations came from an online calculator. I have tried a couple different ones and always end up with similar results. if the tax rate never changes you always end up with the traditional and roth providing an equal retirement income.

So how is one to account for all of these variables and make an informed decision on which is the preferred path if no one can agree on how the calculations should be performed?
It is very simple.

A) Max out your Trad. 401K

B) Put your tax savings into Roth IRAs.

It works 90+% of the time. There are 10% or less of the US household that is millionaires. Why do you need anything else? If you are a millionaire, then, further analysis is needed.

KlangFool

02nz
Posts: 4640
Joined: Wed Feb 21, 2018 3:17 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by 02nz » Fri Jun 26, 2020 11:26 am

Walkure wrote:
Fri Jun 26, 2020 11:19 am
02nz wrote:
Fri Jun 26, 2020 10:41 am
Be careful or you'll get into another marginal vs effective debate! :happy
Wait for it...
yog wrote:
Fri Jun 26, 2020 11:09 am
While the accumulation phase uses the top of the tax bracket for the savings or deduction, the withdrawal phase starts the tax calculation from the bottom, potentially at 0%. This misunderstanding is what trips up many people in the middle brackets today, and it's easy to misuse the online calculators... If we calculate taxes due from a $0 base, we have $2,093 in federal tax, and the effective tax rate is actually %6.7 - much better than the 25% we locked in with the Roth. The traditional ends up giving one 24.5% more spending power annually during retirement ($29,163 vs. $23,422).
And there it is. :oops:

https://www.bogleheads.org/wiki/Traditi ... onceptions
But if one doesn't have any other income - and many retirees do not until SS and RMDs start - then IRA distributions do indeed fill in "from the bottom," starting with the standard deduction, then 10% bracket, 12/15%, and so on. (Actually I think yog assumed single filer; the average rate on that money is even lower for MFJ.)

User avatar
FiveK
Posts: 9226
Joined: Sun Mar 16, 2014 2:43 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by FiveK » Fri Jun 26, 2020 11:30 am

yog wrote:
Fri Jun 26, 2020 11:09 am
While the accumulation phase uses the top of the tax bracket for the savings or deduction, the withdrawal phase starts the tax calculation from the bottom, potentially at 0%. This misunderstanding is what trips up many people....
See the first of two Common misconceptions. As 02nz alluded earlier, this has been discussed ad nauseam. It would be nice if we could start from the bottom on withdrawals from every contribution, but the IRS doesn't allow that. ;)

User avatar
FiveK
Posts: 9226
Joined: Sun Mar 16, 2014 2:43 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by FiveK » Fri Jun 26, 2020 11:37 am

02nz wrote:
Fri Jun 26, 2020 11:26 am
But if one doesn't have any other income....
"Any other income" includes withdrawals based on previous traditional contributions.

If you already have a substantial traditional balance, withdrawals based on that will fill the bottom brackets the same as a pension.

Topic Author
Northern Flicker
Posts: 5780
Joined: Fri Apr 10, 2015 12:29 am

Re: Asymmetric risk in the Roth vs Traditional question

Post by Northern Flicker » Fri Jun 26, 2020 1:40 pm

conservativeinvestor wrote:
Fri Jun 26, 2020 8:52 am
This all seems very confusing.

Aren't the only things that really matter your current tax rate and future tax rate?

Lets try some math:
Salary 100,000 per year
Current tax rate 25%
Retirement tax rate 25%
20 years of contributions
7% rate of return

Traditional Contribution: 18% or 18,000 per year
Roth Contribution: 13.5% or 13,500 per year
BUT: both of these scenarios will take the same amount out of your paycheck so you will receive the same take home pay with either of these choices.

What is the better choice?
Lets do some more calculating:
Traditional ending balance before tax: 781,389
Roth ending balance: 586,042

Traditional ending balance after tax: 586,042
Roth ending balance: 586,042

If your tax bracket never changes it doesn't matter which one you choose.
So all you need to know is: What is your tax bracket now and what will it be when you retire?
Yes, but you do not know your retirement
tax rate, nor how much you will accumulate. Risk is a probabilistic concept that cannot be explained away by making deterministic assumptions about outcomes.
Risk is not a guarantor of return.

Topic Author
Northern Flicker
Posts: 5780
Joined: Fri Apr 10, 2015 12:29 am

Re: Asymmetric risk in the Roth vs Traditional question

Post by Northern Flicker » Fri Jun 26, 2020 1:42 pm

FiveK wrote:
Fri Jun 26, 2020 11:30 am
yog wrote:
Fri Jun 26, 2020 11:09 am
While the accumulation phase uses the top of the tax bracket for the savings or deduction, the withdrawal phase starts the tax calculation from the bottom, potentially at 0%. This misunderstanding is what trips up many people....
See the first of two Common misconceptions. As 02nz alluded earlier, this has been discussed ad nauseam. It would be nice if we could start from the bottom on withdrawals from every contribution, but the IRS doesn't allow that. ;)
Nonetheless, using marginal rate in retirement for making the decision has also been shown to be incorrect in a study by Vanguard.
Risk is not a guarantor of return.

Topic Author
Northern Flicker
Posts: 5780
Joined: Fri Apr 10, 2015 12:29 am

Re: Asymmetric risk in the Roth vs Traditional question

Post by Northern Flicker » Fri Jun 26, 2020 1:50 pm

redmaw wrote:
Fri Jun 26, 2020 7:37 am
Northern Flicker wrote:
Thu Jun 25, 2020 1:47 pm

... Roth assets have a higher after-tax volatility...
Can you explain this assertion? How are roth assets more volatile? I assume the reasoning is 100 goes up to 120, in roth thats a 20% increase and $20, in trad (assuming 22% tax rate) thats 78 aftertax dollars before the increase and 93.6 after the increase see it only changed be $15.6 instead of $20. But 15.6 is still 20% of the after tax ira value, that is the after tax value of both account changed by 20%, the raw dollar amount is different because the traditional account is worth less after tax. Am I missing something?
When trad assets appreciate, your tax drag as a percentage goes up, though as a step function-- it is linear until you cross a tax bracket boundary. The reverse happens when assets depreciate. This dampens volatility of after-tax value on rither side.
Risk is not a guarantor of return.

sc9182
Posts: 331
Joined: Wed Aug 17, 2016 7:43 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by sc9182 » Fri Jun 26, 2020 1:55 pm

aristotelian wrote:
Fri Jun 26, 2020 7:53 am
The way I see it, if it turns out that I am in a higher tax bracket in retirement, it means I "won the game." Being in a higher tax bracket is a good problem to have. The most likely scenario I see this happening is inheritance which gives me income that I am not currently counting on. In that scenario, I won't really regret having chosen traditional. I will lose more of my 401k to tax than anticipated but that is only because I have substantially more assets to draw from.
Even if you receive inheritance - when your Traditional is pretty large; you may disclaim inheritance - and let it pass thru to you younger generation (depending on state laws and GST exclusion limits)
https://corporate.findlaw.com/finance/j ... t-way.html

Show me bigger buckets of money - we will figure best way to make use of it - optimal tax strategy!

By design Roth can’t be/get bigger than Traditional- due to its prepaid/tax structure, and higher susceptibility to SORR risk.

Quoting-Self from Another thread:
“ Strive for achieving higher numbers, yes., realistically possible - but unlikely. Life meanders multiple ways - sequence of returns, marriage, disability, un-employment or under-employment, divorce, death or worse-yet a lawsuit .. :-(

infotrader
Posts: 291
Joined: Tue Feb 28, 2017 2:39 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by infotrader » Fri Jun 26, 2020 2:08 pm

02nz wrote:
Wed Jun 24, 2020 5:43 pm
retired@50 wrote:
Wed Jun 24, 2020 5:35 pm
02nz wrote:
Wed Jun 24, 2020 4:47 pm
I'm sure most newbies asking the Roth vs traditional question have never thought about it; I certainly had not until I read through a lot more Roth vs traditional threads.
The mistake I most often see from young retirement savers is that they feel as though they will be paying higher taxes when they are 35 or 45 years old, so they choose Roth at 25. However, it's the tax rate during retirement that is relevant for comparison, not the tax rate during middle age.

Regards,
To be fair, there's some justification in the thinking you cited, it's not necessarily a mistake. If I'm in the 12% bracket now and expect to be in a higher bracket in a few years because of pay raises (medical residents are kind of an extreme example of this), even with nothing in tax-deferred balances now, I might reasonably decide to contribute to Roth 401k and then later switch to traditional once at 22% or higher, to get more tax diversification. I think for most people the best way to get that diversification is by combining trad'l 401k and Roth IRA (and the point about asymmetric risk applies here - what if those anticipated raises or career advancements don't work out, for example), but starting out with Roth when one's income is at its lowest is not an unreasonable choice, either.
Given the longer time frame for growth, the worst case scenario is probably break even. Also, if you don't contribute to Roth 401k at 25, you probably never will, unless your salary in 35 or 45 is lower.

02nz
Posts: 4640
Joined: Wed Feb 21, 2018 3:17 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by 02nz » Fri Jun 26, 2020 3:49 pm

FiveK wrote:
Fri Jun 26, 2020 11:37 am
02nz wrote:
Fri Jun 26, 2020 11:26 am
But if one doesn't have any other income....
"Any other income" includes withdrawals based on previous traditional contributions.

If you already have a substantial traditional balance, withdrawals based on that will fill the bottom brackets the same as a pension.
Indeed, which is why it's important for posters asking the traditional vs Roth question to give us full details, including size of existing tax-deferred portfolio. (Even better would be the projected size at retirement of existing portfolio plus future employer match which is always traditional, using conservative real, not nominal, projections for returns, but generally newbies don't manage nearly that level of precision.)

User avatar
FiveK
Posts: 9226
Joined: Sun Mar 16, 2014 2:43 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by FiveK » Fri Jun 26, 2020 4:15 pm

02nz wrote:
Fri Jun 26, 2020 3:49 pm
FiveK wrote:
Fri Jun 26, 2020 11:37 am
02nz wrote:
Fri Jun 26, 2020 11:26 am
But if one doesn't have any other income....
"Any other income" includes withdrawals based on previous traditional contributions.

If you already have a substantial traditional balance, withdrawals based on that will fill the bottom brackets the same as a pension.
Indeed, which is why it's important for posters asking the traditional vs Roth question to give us full details, including size of existing tax-deferred portfolio. (Even better would be the projected size at retirement of existing portfolio plus future employer match which is always traditional, using conservative real, not nominal, projections for returns, but generally newbies don't manage nearly that level of precision.)
Agreed on all counts. :sharebeer

User avatar
FiveK
Posts: 9226
Joined: Sun Mar 16, 2014 2:43 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by FiveK » Fri Jun 26, 2020 4:18 pm

Northern Flicker wrote:
Fri Jun 26, 2020 1:42 pm
FiveK wrote:
Fri Jun 26, 2020 11:30 am
yog wrote:
Fri Jun 26, 2020 11:09 am
While the accumulation phase uses the top of the tax bracket for the savings or deduction, the withdrawal phase starts the tax calculation from the bottom, potentially at 0%. This misunderstanding is what trips up many people....
See the first of two Common misconceptions. As 02nz alluded earlier, this has been discussed ad nauseam. It would be nice if we could start from the bottom on withdrawals from every contribution, but the IRS doesn't allow that. ;)
Nonetheless, using marginal rate in retirement for making the decision has also been shown to be incorrect in a study by Vanguard.
If the study in question is the "Break Even Tax Rate (BETR)" study, see Break-even withdrawal rate. The marginal tax rate on the withdrawal amount remains the correct comparison.

If you have a different study in mind, please provide a reference.

NewMoneyMustBeSmart
Posts: 246
Joined: Sat Jun 01, 2019 10:28 pm
Location: Midwest

Re: Asymmetric risk in the Roth vs Traditional question

Post by NewMoneyMustBeSmart » Fri Jun 26, 2020 4:19 pm

02nz wrote:
Wed Jun 24, 2020 4:47 pm
This is exactly right, and the same point I just made in another thread: "I often find that people's projections/assumptions are wildly skewed in favor of Roth ("our deficits are huge!"). They forget that if you choose Roth and turn out to be wrong, it's a costly mistake that can put a dent in your standard of living.
This sort of begs the question that you don't suffer from a problem of plenty.

I am maxing both pre-tax 401k and post-tax Roth and I don't expect this to be a costly mistake.
-- | Few are those who see with their own eyes and feel with their own hearts - Einstein

02nz
Posts: 4640
Joined: Wed Feb 21, 2018 3:17 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by 02nz » Fri Jun 26, 2020 4:35 pm

NewMoneyMustBeSmart wrote:
Fri Jun 26, 2020 4:19 pm
This sort of begs the question that you don't suffer from a problem of plenty.
Isn't that the point of this thread? If traditional turns out to be "wrong," it generally means you "won the game" or have a "problem of plenty." Nice problem to have, but the reverse situation is much more problematic.
NewMoneyMustBeSmart wrote:
Fri Jun 26, 2020 4:19 pm
I am maxing both pre-tax 401k and post-tax Roth and I don't expect this to be a costly mistake.
For one, if you can max a pre-tax 401k and Roth IRA (I assume that's what you meant), you probably don't qualify for a deduction on a traditional IRA, so for the IRA contribution you didn't really face a traditional vs Roth choice. And, note what I wrote below that quote: "I think for most people the best way to get that [tax] diversification is by combining trad'l 401k and Roth IRA." For those say 25-40 years away from retirement, there's too much uncertainty in personal circumstances, tax law, and market returns to to be able to make precise recommendations, so for many the suggestion is indeed "max a traditional 401k and Roth IRA if you can."

Topic Author
Northern Flicker
Posts: 5780
Joined: Fri Apr 10, 2015 12:29 am

Re: Asymmetric risk in the Roth vs Traditional question

Post by Northern Flicker » Fri Jun 26, 2020 5:30 pm

NewMoneyMustBeSmart wrote:
Fri Jun 26, 2020 4:19 pm
02nz wrote:
Wed Jun 24, 2020 4:47 pm
This is exactly right, and the same point I just made in another thread: "I often find that people's projections/assumptions are wildly skewed in favor of Roth ("our deficits are huge!"). They forget that if you choose Roth and turn out to be wrong, it's a costly mistake that can put a dent in your standard of living.
This sort of begs the question that you don't suffer from a problem of plenty.

I am maxing both pre-tax 401k and post-tax Roth and I don't expect this to be a costly mistake.
In fact, the point of this thread is that what you are doing is probably optimal when averaged across all outcomes.
Risk is not a guarantor of return.

Topic Author
Northern Flicker
Posts: 5780
Joined: Fri Apr 10, 2015 12:29 am

Re: Asymmetric risk in the Roth vs Traditional question

Post by Northern Flicker » Fri Jun 26, 2020 6:14 pm

FiveK wrote:
Fri Jun 26, 2020 4:18 pm
Northern Flicker wrote:
Fri Jun 26, 2020 1:42 pm
FiveK wrote:
Fri Jun 26, 2020 11:30 am
yog wrote:
Fri Jun 26, 2020 11:09 am
While the accumulation phase uses the top of the tax bracket for the savings or deduction, the withdrawal phase starts the tax calculation from the bottom, potentially at 0%. This misunderstanding is what trips up many people....
See the first of two Common misconceptions. As 02nz alluded earlier, this has been discussed ad nauseam. It would be nice if we could start from the bottom on withdrawals from every contribution, but the IRS doesn't allow that. ;)
Nonetheless, using marginal rate in retirement for making the decision has also been shown to be incorrect in a study by Vanguard.
If the study in question is the "Break Even Tax Rate (BETR)" study, see Break-even withdrawal rate. The marginal tax rate on the withdrawal amount remains the correct comparison.

If you have a different study in mind, please provide a reference.
That is the study. I don't agree with the study or with the wiki but it will not be fruitful to discuss.
Risk is not a guarantor of return.

User avatar
FiveK
Posts: 9226
Joined: Sun Mar 16, 2014 2:43 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by FiveK » Fri Jun 26, 2020 6:40 pm

Northern Flicker wrote:
Fri Jun 26, 2020 6:14 pm
FiveK wrote:
Fri Jun 26, 2020 4:18 pm
Northern Flicker wrote:
Fri Jun 26, 2020 1:42 pm
Nonetheless, using marginal rate in retirement for making the decision has also been shown to be incorrect in a study by Vanguard.
If the study in question is the "Break Even Tax Rate (BETR)" study, see Break-even withdrawal rate. The marginal tax rate on the withdrawal amount remains the correct comparison.
That is the study. I don't agree with the study or with the wiki but it will not be fruitful to discuss.
OK.

For anyone else interested, the Vanguard study is https://personal.vanguard.com/pdf/ISGBETR.pdf. On p. 2 one finds "You should consider your future marginal tax rate...when thinking about Roth conversions."

columbia
Posts: 2662
Joined: Tue Aug 27, 2013 5:30 am

Re: Asymmetric risk in the Roth vs Traditional question

Post by columbia » Fri Jun 26, 2020 7:53 pm

Since most people have a 401k-style plan, defaulting to Roth seems reasonable: you can’t know your future rates, so split the difference between Roth vs 401k/IRA/etc.
If you leave your head in the sand for too long, you might get run over by a Jeep.

02nz
Posts: 4640
Joined: Wed Feb 21, 2018 3:17 pm

Re: Asymmetric risk in the Roth vs Traditional question

Post by 02nz » Fri Jun 26, 2020 7:55 pm

columbia wrote:
Fri Jun 26, 2020 7:53 pm
Since most people have a 401k-style plan, defaulting to Roth seems reasonable: you can’t know your future rates, so split the difference between Roth vs 401k/IRA/etc.
Do you mean using both a traditional 401k AND Roth IRA? Roth can apply to 401k or IRA, so "split the difference between Roth vs 401k/IRA/etc." is unclear.

Post Reply