I am helping a friend, and want to confirm so that I advise them accurately.

I have searched for things like "tax drag" on this website which loosely covers aspects of this situation. Please send a forum link if this specific question has already been covered. I still wanted to ask for this specific case in the event there were any unique considerations.

Situation:
They are mid-30s year old now. (so, say roughly 30-40 years until withdrawal of a contributed dollar)
22% bracket now. Likely in meat of 24% bracket in several years and for most of career. Due to a change in the nature of their compensation, they will have no access to a workplace retirement account for next several years despite having an identical job.
Assume person wants to contribute the maximum every year to their retirement account.
Workplace has traditional retirement account with great low-cost choices.

The question is: how much "extra" would the employer need to compensate them per year to account for the lost access to tax-advantaged retirement space?
Or, put another way (as a hypothetical), how much should this person be willing to pay for access to a workplace retirement account where they could contribute $19,500 of traditional tax-advantaged dollars.
(ignore any matching contributions, just value of the "space")

My guess was that I would value the access to retirement space at around 2.5k-5k if I did the math.

Then I tried to do the math.
Assumed 7.5% IRR.
2% dividend and average 20% tax on dividends, and
15% LTCG.
In 30 years, 19.5k turns into about 124k in a taxable (after accounting for LTCG and dividend drag).
In tax-advantaged account, amount was roughly 30k more than 124k taxable account in 2050 dollars.
(i assumed marginal rate in was identical to effective rate out, which should be generous, if anything)
The present value of this 30k difference (2050 dollars) is in the 10-15k range depending on inflation assumptions.
10-15k was much more than I thought it would be. Did I screw up?
No looking to split hairs here about 7 dollars and realize there are a lot of assumptions about future things that impact this (rate of return, tax rates, etc).

Returning to the specific question: How much (roughly) should this person need to get compensated to make up for the loss of $19.5k of tax-advantaged space?

I tried to duplicate this and I came out pretty close ignoring the taxes paid at the beginning and paying capital gains tax only once. Is that how you did it? In a 401k, I got $40,000 more in 30 years, but $30,000 more in 29 years.

Another issue I see everywhere (not here) is converting to present value using zero percent interest. The truth is, we wouldn't really accept realistic net present value calculations. We don't really emotionally accept that the cost of capital is zero. But based on that, I think $10,000 is an okay number. I had to use about 5% to get from 40k down to 10.

Last edited by firebirdparts on Thu Sep 24, 2020 1:45 pm, edited 1 time in total.

firebirdparts wrote: ↑Thu Sep 24, 2020 1:18 pm
I tried to duplicate this and I came out pretty close ignoring the taxes paid at the beginning and paying capital gains tax only once. Is that how you did it? In a 401k, I got $40,000 more in 30 years, but $30,000 more in 29 years.

Another issue I see everywhere (not here) is converting to present value using zero percent interest. The truth is, we wouldn't really accept realistic net present value calculations. We don't really emotionally accept that the cost of capital is zero. But based on that, I think $10,000 is an okay number.

Yes - I just assumed withdraw everything 30 years in the future (at 15% LTCG rate).

A related thought is that an increased rate of return will enhance the "cost" of the lost retirement access. But, if one subscribes to the idea that real return might be somewhat more constant than total returns, than some if this extra cost shakes out when doing NPV calculation.

I am a bit confused how you got 40k more in 30 years, but 30k more in 29 years. Only 1 year more drove a 10k difference?

firebirdparts wrote: ↑Thu Sep 24, 2020 1:18 pm
I tried to duplicate this and I came out pretty close ignoring the taxes paid at the beginning and paying capital gains tax only once. Is that how you did it? In a 401k, I got $40,000 more in 30 years, but $30,000 more in 29 years.

Another issue I see everywhere (not here) is converting to present value using zero percent interest. The truth is, we wouldn't really accept realistic net present value calculations. We don't really emotionally accept that the cost of capital is zero. But based on that, I think $10,000 is an okay number.

Yes - I just assumed withdraw everything 30 years in the future (at 15% LTCG rate).

A related thought is that an increased rate of return will enhance the "cost" of the lost retirement access. But, if one subscribes to the idea that real return might be somewhat more constant than total returns, than some if this extra cost shakes out when doing NPV calculation.

I am a bit confused how you got 40k more in 30 years, but 30k more in 29 years. Only 1 year more drove a 10k difference?

Sure. $160k X 1.075 is over $170k. In taxable, I got $129k but I thought that was pretty close. There are some minor differences that arise from using last year's total to calculate this year's dividends and those sorts of things.

firebirdparts wrote: ↑Thu Sep 24, 2020 1:18 pm
I tried to duplicate this and I came out pretty close ignoring the taxes paid at the beginning and paying capital gains tax only once. Is that how you did it? In a 401k, I got $40,000 more in 30 years, but $30,000 more in 29 years.

Another issue I see everywhere (not here) is converting to present value using zero percent interest. The truth is, we wouldn't really accept realistic net present value calculations. We don't really emotionally accept that the cost of capital is zero. But based on that, I think $10,000 is an okay number.

Yes - I just assumed withdraw everything 30 years in the future (at 15% LTCG rate).

A related thought is that an increased rate of return will enhance the "cost" of the lost retirement access. But, if one subscribes to the idea that real return might be somewhat more constant than total returns, than some if this extra cost shakes out when doing NPV calculation.

I am a bit confused how you got 40k more in 30 years, but 30k more in 29 years. Only 1 year more drove a 10k difference?

Sure. $160k X 1.075 is over $170k. In taxable, I got $129k but I thought that was pretty close. There are some minor differences that arise from using last year's total to calculate this year's dividends and those sorts of things.

But the question is the RELATIVE cost between taxable and retirement? So in my mind it would be the extra LTCG and dividend for the extra year, and not the entire accounts increase in "value" (taxes not withstanding) due to investment return.

Using rows 131-147 on the personal finance toolbox "Misc. calcs' tab for taxable growth suggests $137,185 after 30 years, and $128,241 after 29.

The straightforward $19.5K * (1 + i)^n calculations for i = 0.075 and n = 30 and 29 gives $170,722 and $158,811 respectively. Differences of $33,536 for 30 years and $30,570 for 29 years.

This is a reasonable comparison between taxable and Roth accounts, because both start with after-tax money.

Comparing taxable vs. traditional requires the additional estimate of current vs. future marginal tax rates on ordinary income.

FiveK wrote: ↑Thu Sep 24, 2020 2:01 pm
Using rows 131-147 on the personal finance toolbox "Misc. calcs' tab for taxable growth suggests $137,185 after 30 years, and $128,241 after 29.

The straightforward $19.5K * (1 + i)^n calculations for i = 0.075 and n = 30 and 29 gives $170,722 and $158,811 respectively. Differences of $33,536 for 30 years and $30,570 for 29 years.

This is a reasonable comparison between taxable and Roth accounts, because both start with after-tax money.

Comparing taxable vs. traditional requires the additional estimate of current vs. future marginal tax rates on ordinary income.

Thanks. I did not know about that spreadsheet!
I basically made the crappy 20 minute home-brew version of it for my very specific case.

For thinking about this, I kind of like comparing to Roth accounts just for the sake of simplicity. (traditional only available at work, but for comparison to taxable using taxed roth dollars in conveinent for high-level analysis).
Of course, I think marginal rate in = effective rate out makes a traditional identical to a roth (although we can argue contribution limits are not identical due to pre/post tax nature!)

In general, I'm OK not getting in the weeds about likely future tax rates and what withdrawal amounts are likely to be and effective tax rates in retirement. I realize they are drivers of value/cost here.

blastoff wrote: ↑Thu Sep 24, 2020 2:08 pm
Thanks. I did not know about that spreadsheet!

I mostly use it as a double check on our own home-grown tax spreadsheet, but it does have other bells and whistles such as this.

For thinking about this, I kind of like comparing to Roth accounts just for the sake of simplicity. ...
In general, I'm OK not getting in the weeds about likely future tax rates....

I instead used 4% real rate. Over 30 years that is a factor of 3.25.
19500 times 3.25 times.76 = $48k for traditional scenario.
19500 times.76 times 3.25 times.85 = $41k for taxable.

I ignored annual dividend impact and assumed taxable all came out at end as capital gains. So the difference is $7k for one year of contributions. That could go up or down materially if you used a different real compounding rate.

Question, is the new opportunity contnractor/1099? If so friend could do a solo 401k.

JBTX wrote: ↑Thu Sep 24, 2020 2:28 pm
I instead used 4% real rate. Over 30 years that is a factor of 3.25.
19500 times 3.25 times.76 = $48k for traditional scenario.
19500 times.76 times 3.25 times.85 = $41k for taxable.

I ignored annual dividend impact and assumed taxable all came out at end as capital gains. So the difference is $7k for one year of contributions. That could go up or down materially if you used a different real compounding rate.

Question, is the new opportunity contnractor/1099? If so friend could do a solo 401k.

No. Welcome to the stupid world of academics that goes something like this:

Friend: "Instead of you having to pay me to do the job with your money, I worked hard and wrote a lot of words down on a piece of paper and got money to pay my salary. So now, you'll just get money pay me with that I went out and got for you and you don't have to pay me anymore with your money."
Employer: "Great! - in reward you giving us money to pay your salary so we don't have to pay you, we're going to take your benefits"

JBTX wrote: ↑Thu Sep 24, 2020 2:28 pm
I instead used 4% real rate. Over 30 years that is a factor of 3.25.
19500 times 3.25 times.76 = $48k for traditional scenario.
19500 times.76 times 3.25 times.85 = $41k for taxable.

I ignored annual dividend impact and assumed taxable all came out at end as capital gains. So the difference is $7k for one year of contributions. That could go up or down materially if you used a different real compounding rate.

Question, is the new opportunity contnractor/1099? If so friend could do a solo 401k.

No. Welcome to the stupid world of academics that goes something like this:

Friend: "Instead of you having to pay me to do the job with your money, I worked hard and wrote a lot of words down on a piece of paper and got money to pay my salary. So now, you'll just get money pay me with that I went out and got for you and you don't have to pay me anymore with your money."
Employer: "Great! - in reward you giving us money to pay your salary so we don't have to pay you, we're going to take your benefits"

JBTX wrote: ↑Thu Sep 24, 2020 2:28 pm
I instead used 4% real rate. Over 30 years that is a factor of 3.25.
19500 times 3.25 times.76 = $48k for traditional scenario.
19500 times.76 times 3.25 times.85 = $41k for taxable.

I ignored annual dividend impact and assumed taxable all came out at end as capital gains. So the difference is $7k for one year of contributions. That could go up or down materially if you used a different real compounding rate.

Question, is the new opportunity contnractor/1099? If so friend could do a solo 401k.

No. Welcome to the stupid world of academics that goes something like this:

Friend: "Instead of you having to pay me to do the job with your money, I worked hard and wrote a lot of words down on a piece of paper and got money to pay my salary. So now, you'll just get money pay me with that I went out and got for you and you don't have to pay me anymore with your money."
Employer: "Great! - in reward you giving us money to pay your salary so we don't have to pay you, we're going to take your benefits"

I have no idea what that means, but that's OK.

If you get a grant to pay your whole salary you are no longer an employee with retirement benefits.
(they usually give you some similar-ish health insurance plan)

In an effort to get back to the initial question (that I partially derailed!):

What would people see as the cost of not having the retirement account "space"?
10k? 7k? More? Less?
I think it's about 10k on the mid-low end with reasonable assumptions.

I realize that in addition to the rate of return assumptions, the difference between 40 or 30 years could have a large impact on the cost.

I'm looking for a fair but very defensible number that might fall in the 25-50% range (low to mid end)

blastoff wrote: ↑Thu Sep 24, 2020 2:48 pm
In an effort to get back to the initial question (that I partially derailed!):

What would people see as the cost of not having the retirement account "space"?
10k? 7k? More? Less?
I think it's about 10k on the mid-low end with reasonable assumptions.

I realize that in addition to the rate of return assumptions, the difference between 40 or 30 years could have a large impact on the cost.

I'm looking for a fair but very defensible number that might fall in the 25-50% range (low to mid end)

Did you see the first part of my post? I came up with $7k per year. I'd say between $5k and $10k depending on your assumed future real interest rate.

blastoff wrote: ↑Thu Sep 24, 2020 2:48 pm
In an effort to get back to the initial question (that I partially derailed!):

What would people see as the cost of not having the retirement account "space"?
10k? 7k? More? Less?
I think it's about 10k on the mid-low end with reasonable assumptions.

I realize that in addition to the rate of return assumptions, the difference between 40 or 30 years could have a large impact on the cost.

I'm looking for a fair but very defensible number that might fall in the 25-50% range (low to mid end)

Did you see the first part of my post? I came up with $7k per year. I'd say between $5k and $10k depending on your assumed future real interest rate.

Yes thanks! Just wondering if that is a consensus number or others would push back and think longer time horizon or additional costs should be considered.

I got about 10k on the low end. I think you ignored dividend impact which is far from an insignificant driver of cost (maybe ~2/5 of total).

I realize there is no right answer, but amount was higher than I thought so wanted to double check my "work" with people here.

blastoff wrote: ↑Thu Sep 24, 2020 2:58 pm
I realize there is no right answer, but amount was higher than I thought so wanted to double check my "work" with people here.

It's a number, and it's defensible.

The question now becomes "what does the school administration think is the correct number?" - good luck with that!

blastoff wrote: ↑Thu Sep 24, 2020 2:58 pm
I realize there is no right answer, but amount was higher than I thought so wanted to double check my "work" with people here.

It's a number, and it's defensible.

The question now becomes "what does the school administration think is the correct number?" - good luck with that!

Ha - I was in academics awhile so normally I would agree with you 100% and expect that this would be a difficult battle unlikely to lead to a good outcome despite the obvious mathematics of it all. However, for several reasons I think a good chance this will be resolved reasonably. Namely, I know there is a lot of money available (this is a rounding error to them) and there exists some precedent.

blastoff wrote: ↑Thu Sep 24, 2020 2:48 pm
In an effort to get back to the initial question (that I partially derailed!):

What would people see as the cost of not having the retirement account "space"?
10k? 7k? More? Less?
I think it's about 10k on the mid-low end with reasonable assumptions.

I realize that in addition to the rate of return assumptions, the difference between 40 or 30 years could have a large impact on the cost.

I'm looking for a fair but very defensible number that might fall in the 25-50% range (low to mid end)

Did you see the first part of my post? I came up with $7k per year. I'd say between $5k and $10k depending on your assumed future real interest rate.

Yes thanks! Just wondering if that is a consensus number or others would push back and think longer time horizon or additional costs should be considered.

I got about 10k on the low end. I think you ignored dividend impact which is far from an insignificant driver of cost (maybe ~2/5 of total).

I realize there is no right answer, but amount was higher than I thought so wanted to double check my "work" with people here.

If dividends are qualified then the tax rate is the same. They are just taxed sooner and added to the basis for less capital gains. You lose a bit of compounding on the taxable side but I doubt the answer will be materially different.

blastoff wrote: ↑Thu Sep 24, 2020 2:48 pm
In an effort to get back to the initial question (that I partially derailed!):

What would people see as the cost of not having the retirement account "space"?
10k? 7k? More? Less?
I think it's about 10k on the mid-low end with reasonable assumptions.

I realize that in addition to the rate of return assumptions, the difference between 40 or 30 years could have a large impact on the cost.

I'm looking for a fair but very defensible number that might fall in the 25-50% range (low to mid end)

Did you see the first part of my post? I came up with $7k per year. I'd say between $5k and $10k depending on your assumed future real interest rate.

Yes thanks! Just wondering if that is a consensus number or others would push back and think longer time horizon or additional costs should be considered.

I got about 10k on the low end. I think you ignored dividend impact which is far from an insignificant driver of cost (maybe ~2/5 of total).

I realize there is no right answer, but amount was higher than I thought so wanted to double check my "work" with people here.

If dividends are qualified then the tax rate is the same. They are just taxed sooner and added to the basis for less capital gains. You lose a bit of compounding on the taxable side but I doubt the answer will be materially different.

Ah- makes sense. Let me double check my math. May have not accounted for basis change.