Kitces on Roth Conversions vs. Harvesting Capital Gains

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Topic Author
petulant
Posts: 1901
Joined: Thu Sep 22, 2016 1:09 pm

Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by petulant »

Michael Kitces has a new blog post that does a great job illustrating issues related to Roth conversions and harvesting capital gains. Here's a link:

https://www.kitces.com/blog/navigating- ... nversions/

I especially appreciate Kitces drawing attention to the reality that any opportunity to do Roth conversions in the 12% bracket is matched to an opportunity to realize capital gains in the 0% LTCG bracket. He explains thoroughly how low income tax brackets and the 0% capital gains bracket have a shared space that can only be used for one or the other or a finite combination of both. He shows that, at the very least, investors should realize ordinary income to fill the standard deduction or other deductions first, then consider capital gains vs. ordinary income (like Roth conversions) in higher brackets.

Unfortunately, Kitces appears to assume that the default priority should be to realize capital gains tax at the 0% rate rather than recognize ordinary income tax at the 10 and 12% rates. Kitces appears focused on the fact that capital gains taxed at 0% have a 15% differential from the next bracket, while ordinary income only has a 10% differential between the 12% and 22% brackets. While this may be true for some investors, it is problematic for four reasons.

One, investors can choose to defer capital gains as long as they wish, but they can't defer IRA income forever. Investors thus have much more control over capital gains once they begin taking social security, but RMDs can cause taxability in social security regardless of investor wishes. Realizing as much ordinary income as possible before social security begins can therefore reduce social security taxation in the majority of years while still leaving the investor in control of when to realize capital gains. While I haven't quantified it--and it would depend on the investor--I believe this means the savings from realizing ordinary income through a Roth conversion before SS is more than the 10% bracket differential between 12% and 22%. Kitces does not mention this directly.

Two, if investors do need additional money beyond retirement account withdrawals, social security, and other sources in the future, capital gains even taxed at the 15% rate are more efficient since any realization of capital gains comes with a return of basis. In other words, to get $100 out of the IRA, the investor has to pay taxes on the whole amount. To get $100 out of a taxable account, the investor will only pay taxes on the gains. Further, the investor can target the most desirable ratio based on tax lots available in the account. Capital gains assets are more efficient for obtaining a specific amount of cash with minimum taxation while affording significant flexibility. Kitces does not mention this; in fact, his example assumes a zero basis.

Three, and this is the most important one, assets with capital gains enjoy the step-up basis at death, while IRA balances do not. An investor really will be wasting time recognizing gains at 0% if the investor could have just realized ordinary income from the IRA while allowing the step-up basis to eliminate capital gains taxes when leaving assets to heirs. Fortunately, Kitces lists this as a "mitigating factor," but I think it deserves to be at the top of the considerations, not buried at the bottom.

Fourth, recognizing ordinary income through a Roth conversion allows all future growth to be free of taxes, while recognizing capital gains does not. In other words, if I perform a Roth conversion to move $100 from the IRA to the Roth IRA, I no longer have any tax liability from future growth in the asset. If it goes up 7%, I get the full 7%. For those who like the "government loan" analogy for IRAs, I have fully and finally paid off the loan. Not so for realizing capital gains. That just pays the tax bill for a specific nominal dollar amount, but it doesn't relieve the investor of taxes on additional growth in the future. While the relationship between this issue and risk can be complex, I nevertheless think getting the money in the Roth account can create a net present value of savings from avoiding taxes on future growth compared to the taxable account.

Overall, Kitces does a great job explaining these issues in the article. I just think he makes a naive assumption that the 15% differential between capital gains taxes trumps the 10% differential for ordinary income. The default, in my opinion, really should be to recognize ordinary income up to the top of the 12% bracket in years before SS starts, and that should only really shift for more specific situations. Beyond the 12% bracket, obviously, it gets more complex as the bracket differentials and IRMAA issues require more specific planning.
livesoft
Posts: 73338
Joined: Thu Mar 01, 2007 8:00 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by livesoft »

Thanks for this. I think you should send your comment to Kitces or post it as a response to his blog or e-mail it to him.

I also prefer to do Roth conversions unless I need some money to pay expenses.
Wiki This signature message sponsored by sscritic: Learn to fish.
02nz
Posts: 5555
Joined: Wed Feb 21, 2018 3:17 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by 02nz »

Good points. Related to this, I think in many experts' advice there's often a bias toward liquidating taxable accounts first and letting tax-advantaged grow (here's another example: https://www.morningstar.com/articles/84 ... ithdrawals), when in many cases it's actually better to first draw down/convert tax-deferred balances to manage taxes/RMDs, and leave the taxable for later in the hope of a step-up in basis (keeping in mind tax laws can and do change).
User avatar
FiveK
Posts: 10290
Joined: Sun Mar 16, 2014 2:43 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FiveK »

The '0% LTCG or t->R' tab in the personal finance toolbox spreadsheet may be useful for comparison:
This addresses the situation of being in a low bracket (12% or less) this year and having two possibilities:
1) Tax Gain Harvest: incur Long Term Capital gains but pay $0 federal tax
2) Convert traditional to Roth at 12% federal, with the expectation of being in a higher tax bracket after retirement
From the limited use I've made of that tool, it seems Roth conversions are preferred over tax gain harvesting when
- the future marginal tax rate is assumed to be higher than 12%, and
- money is expected to remain invested for more than a few years.
marcopolo
Posts: 3468
Joined: Sat Dec 03, 2016 10:22 am

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by marcopolo »

petulant wrote: Wed Jul 22, 2020 9:47 am Michael Kitces has a new blog post that does a great job illustrating issues related to Roth conversions and harvesting capital gains. Here's a link:

https://www.kitces.com/blog/navigating- ... nversions/

I especially appreciate Kitces drawing attention to the reality that any opportunity to do Roth conversions in the 12% bracket is matched to an opportunity to realize capital gains in the 0% LTCG bracket. He explains thoroughly how low income tax brackets and the 0% capital gains bracket have a shared space that can only be used for one or the other or a finite combination of both. He shows that, at the very least, investors should realize ordinary income to fill the standard deduction or other deductions first, then consider capital gains vs. ordinary income (like Roth conversions) in higher brackets.

Unfortunately, Kitces appears to assume that the default priority should be to realize capital gains tax at the 0% rate rather than recognize ordinary income tax at the 10 and 12% rates. Kitces appears focused on the fact that capital gains taxed at 0% have a 15% differential from the next bracket, while ordinary income only has a 10% differential between the 12% and 22% brackets. While this may be true for some investors, it is problematic for four reasons.

One, investors can choose to defer capital gains as long as they wish, but they can't defer IRA income forever. Investors thus have much more control over capital gains once they begin taking social security, but RMDs can cause taxability in social security regardless of investor wishes. Realizing as much ordinary income as possible before social security begins can therefore reduce social security taxation in the majority of years while still leaving the investor in control of when to realize capital gains. While I haven't quantified it--and it would depend on the investor--I believe this means the savings from realizing ordinary income through a Roth conversion before SS is more than the 10% bracket differential between 12% and 22%. Kitces does not mention this directly.

Two, if investors do need additional money beyond retirement account withdrawals, social security, and other sources in the future, capital gains even taxed at the 15% rate are more efficient since any realization of capital gains comes with a return of basis. In other words, to get $100 out of the IRA, the investor has to pay taxes on the whole amount. To get $100 out of a taxable account, the investor will only pay taxes on the gains. Further, the investor can target the most desirable ratio based on tax lots available in the account. Capital gains assets are more efficient for obtaining a specific amount of cash with minimum taxation while affording significant flexibility. Kitces does not mention this; in fact, his example assumes a zero basis.

Three, and this is the most important one, assets with capital gains enjoy the step-up basis at death, while IRA balances do not. An investor really will be wasting time recognizing gains at 0% if the investor could have just realized ordinary income from the IRA while allowing the step-up basis to eliminate capital gains taxes when leaving assets to heirs. Fortunately, Kitces lists this as a "mitigating factor," but I think it deserves to be at the top of the considerations, not buried at the bottom.

Fourth, recognizing ordinary income through a Roth conversion allows all future growth to be free of taxes, while recognizing capital gains does not. In other words, if I perform a Roth conversion to move $100 from the IRA to the Roth IRA, I no longer have any tax liability from future growth in the asset. If it goes up 7%, I get the full 7%. For those who like the "government loan" analogy for IRAs, I have fully and finally paid off the loan. Not so for realizing capital gains. That just pays the tax bill for a specific nominal dollar amount, but it doesn't relieve the investor of taxes on additional growth in the future. While the relationship between this issue and risk can be complex, I nevertheless think getting the money in the Roth account can create a net present value of savings from avoiding taxes on future growth compared to the taxable account.

Overall, Kitces does a great job explaining these issues in the article. I just think he makes a naive assumption that the 15% differential between capital gains taxes trumps the 10% differential for ordinary income. The default, in my opinion, really should be to recognize ordinary income up to the top of the 12% bracket in years before SS starts, and that should only really shift for more specific situations. Beyond the 12% bracket, obviously, it gets more complex as the bracket differentials and IRMAA issues require more specific planning.
Nice analysis.

Do you think any of that changes (or if either is even worthwhile) if one is an early retiree withdrawing in 12% bracket, but also taking advantage of ACA Tax credits?
This raises the effective rates of either by 9.8%. I want to be doing Roth Conversions, but am still trying to convince myself it makes sense.
Once in a while you get shown the light, in the strangest of places if you look at it right.
H-Town
Posts: 2868
Joined: Sun Feb 26, 2017 2:08 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by H-Town »

Thanks for the analysis!

I agree that Roth conversion would have more priority than harvesting capital gain. You could harvest capital gain at 0% LTCG after you fill up the standard/itemized deduction space for Roth conversion.

Minimizing tax = more money in your pocket :sharebeer
User avatar
FIREchief
Posts: 5309
Joined: Fri Aug 19, 2016 6:40 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FIREchief »

That's a good article. It's rare (and difficult) that somebody tries to cover the broader picture including 0% LTCG tax gain harvesting, Roth conversions, IRMAA, etc. in a single article. That said, he only has one example of MFJ, and it doesn't get into the much more difficult "what if" planning for one of the spouses dying early, where the surviving spouse's marginal rate may spike but they may pay no more in LTCG taxes.

It's timely for me because I've recently been running some comparisons to attempt to prioritize tax gain harvesting vs. Roth conversions. Unfortunately, we really have no idea if we'll really revert to the pre-2018 federal tax structure after 2025 (as is currently the law). I see the variables out of Washington as harder to deal with than market return forecasts or life expectancies (which are all of course unknowable). I can plan as if DW and I will both live to 99 and require collectively ten years of LTC. I can also adopt a conservative market return forecast. Both of those will add "safety" to any plans. I can't really do something similar with the tax code in the future.
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.
aristotelian
Posts: 8055
Joined: Wed Jan 11, 2017 8:05 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by aristotelian »

Good summary and even better takedown!

Only thing I would add, when you have a complex question and you aren't sure which is best, you can always do a little of each.
deltaneutral83
Posts: 1703
Joined: Tue Mar 07, 2017 4:25 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by deltaneutral83 »

So on the surface mathematically, the spread between LTCG at 0 and 15% is greater and gives the appearance of being the better option (for those retired and not yet on SS) against Roth conversions up to the 12% bucket (because the next bucket is 22% creating only 10% spread) but in the end Roth conversions up to the 12% bucket (assuming MFJ) is more than likely to be the better option over 0% LTCG for a host of other reasons?
User avatar
FiveK
Posts: 10290
Joined: Sun Mar 16, 2014 2:43 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FiveK »

deltaneutral83 wrote: Wed Jul 22, 2020 3:37 pm So on the surface mathematically, the spread between LTCG at 0 and 15% is greater and gives the appearance of being the better option (for those retired and not yet on SS) against Roth conversions up to the 12% bucket (because the next bucket is 22% creating only 10% spread) but in the end Roth conversions up to the 12% bucket (assuming MFJ) is more than likely to be the better option over 0% LTCG for a host of other reasons?
You could try running your own assumptions through the Excel tool mentioned in this post earlier in the thread.

As with many things, probably "it depends..." on one's specific circumstances.
User avatar
LilyFleur
Posts: 1500
Joined: Fri Mar 02, 2018 10:36 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by LilyFleur »

aristotelian wrote: Wed Jul 22, 2020 3:18 pm Good summary and even better takedown!

Only thing I would add, when you have a complex question and you aren't sure which is best, you can always do a little of each.
Last year, I posted here asking if I should do a Roth conversion (hadn't done any yet), and this advice was the best I got. (It might have been you, I don't remember.) It's a complex decision with a lot of moving parts. And, as we saw recently with stretch IRAs going away, tax law can change.
User avatar
FIREchief
Posts: 5309
Joined: Fri Aug 19, 2016 6:40 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FIREchief »

aristotelian wrote: Wed Jul 22, 2020 3:18 pm Good summary and even better takedown!

Only thing I would add, when you have a complex question and you aren't sure which is best, you can always do a little of each.
Yep. I've been looking into an alternating years scheme. I'm not at IRMAA age (63) and won't claim SS until 70 (DW will at FRA). So, I have a few years during which I can fully control/manipulate my income amounts and sources (other than after-tax dividend and CG distributions). I believe that in my situation I will benefit from using year "a" to execute a Roth conversion up to the remainder of my standard deduction and then sell appreciated equities up to an AGI of $80,000 (the current top of the 0% LTCG "bracket"). This will be a zero tax liability year. I'll retain any cash needed for year "b" to perform a Roth conversion up to the top of the 22% or 24% tax bracket without selling any appreciated equities. Wash, rinse, repeat. Year "c" will be just like year "a." Year "d" will be just like year "b," etc.

The numbers change a bit as DW and I reach age 63 and IRMAA enters the equation. I think the article suggested that IRMAA "only" adds a 1% equivalent tax surcharge, but the current numbers are 1.58%, 1.93% and 1.93% 3.82%, 4.72 and 4.71 for the first three IRMAA tiers. Possibly Not prohibitive, but worth consideration.

edited to correct tax surcharge percentages
Last edited by FIREchief on Wed Jul 22, 2020 7:30 pm, edited 2 times in total.
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.
kaneohe
Posts: 6688
Joined: Mon Sep 22, 2008 12:38 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by kaneohe »

FIREchief wrote: Wed Jul 22, 2020 4:13 pm ........................................................................................

The numbers change a bit as DW and I reach age 63 and IRMAA enters the equation. I think the article suggested that IRMAA "only" adds a 1% equivalent tax surcharge, but the current numbers are 1.58%, 1.93% and 1.93% for the first three IRMAA tiers. Not prohibitive, but worth consideration.
Can you show how that is calculated...........I seem to be getting more like 5%........
User avatar
FiveK
Posts: 10290
Joined: Sun Mar 16, 2014 2:43 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FiveK »

kaneohe wrote: Wed Jul 22, 2020 4:33 pm Can you show how that is calculated...........I seem to be getting more like 5%........
+1

At best it would be 3.8%, 4.7%, and 4.7% to go over the $87K, $109K, and $136K tiers respectively for Medicare B and D - and that's if one goes all the way up exactly to the next tier.

E.g.,
12*((202.4+12.2)-(144.6+0.00))/(109000-87000) = 3.8%
12*((289.2+31.5)-(202.4+12.2))/(136000-109000) = 4.7%
User avatar
ray.james
Posts: 1378
Joined: Tue Jul 19, 2011 4:08 am

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by ray.james »

LilyFleur wrote: Wed Jul 22, 2020 4:10 pm
aristotelian wrote: Wed Jul 22, 2020 3:18 pm Good summary and even better takedown!

Only thing I would add, when you have a complex question and you aren't sure which is best, you can always do a little of each.
Last year, I posted here asking if I should do a Roth conversion (hadn't done any yet), and this advice was the best I got. (It might have been you, I don't remember.) It's a complex decision with a lot of moving parts. And, as we saw recently with stretch IRAs going away, tax law can change.
I think this is also a factor to consider. Changes to tax code for taxable account are more probable than to Roth. There have been 7-8 tax plans thrown around in this election cycle, plus house senate proposed plans in past 2 years. A common sum denominator gives a decent picture of probable future changes.

Great post OP.
When in doubt, http://www.bogleheads.org/forum/viewtopic.php?f=1&t=79939
shess
Posts: 663
Joined: Wed May 17, 2017 12:02 am

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by shess »

petulant wrote: Wed Jul 22, 2020 9:47 am Michael Kitces has a new blog post that does a great job illustrating issues related to Roth conversions and harvesting capital gains. Here's a link:

https://www.kitces.com/blog/navigating- ... nversions/

I especially appreciate Kitces drawing attention to the reality that any opportunity to do Roth conversions in the 12% bracket is matched to an opportunity to realize capital gains in the 0% LTCG bracket. He explains thoroughly how low income tax brackets and the 0% capital gains bracket have a shared space that can only be used for one or the other or a finite combination of both. He shows that, at the very least, investors should realize ordinary income to fill the standard deduction or other deductions first, then consider capital gains vs. ordinary income (like Roth conversions) in higher brackets.
Thanks for the link!

Last year, I did this analysis for us. I'm 51, my spouse is 49, so we have plenty of time before RMDs. We have some substantial positions with significant LTCG, so initially I was strongly tilted towards harvesting those LTCG. BUT, after working things back and forth, it really felt like Roth conversions were worth more over the entire lifecycle, especially if one of us were to die earlier, forcing distributions to accelerate.

But this year's shenanigans demonstrated to me that I was pushing my luck on our asset allocation. I was making an argument against rebalancing, because I was greedy (basically). So I've made some sales to take advantage of the market and get things in order, and now I'm thinking I should probably run with that to some extent. Still working the numbers, but it feels like swapping some of our taxable holdings from equities to fixed-income, and vice versa in tax-deferred, makes sense insofar as that taxable fixed-income is then drawn down over the next few years to support living expenses. That in turn lets us increase Roth conversions over the next few years, because we won't have to balance against LTCG sales.

I'm almost tempted to build a tax-estimator module so that I can just write programs to run scenarios and generate graphs and heatmaps. It feels like the spreadsheet-based systems end up having self-referential issues, at least in Google Sheets. It would be interesting if a constraint on the tax code would be a requirement for the IRS to publish a code module to implement things :-).
User avatar
FiveK
Posts: 10290
Joined: Sun Mar 16, 2014 2:43 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FiveK »

shess wrote: Wed Jul 22, 2020 5:07 pm I'm almost tempted to build a tax-estimator module so that I can just write programs to run scenarios and generate graphs and heatmaps. It feels like the spreadsheet-based systems end up having self-referential issues, at least in Google Sheets.
From what I understand, Excel just works better than Sheets, at least for the automated charting the personal finance toolbox does. E.g., see Any Google Sheets knowledge?.

The Excel1040 calculator doesn't generate charts, so it may work ok in Sheets.
Topic Author
petulant
Posts: 1901
Joined: Thu Sep 22, 2016 1:09 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by petulant »

marcopolo wrote: Wed Jul 22, 2020 1:47 pm
petulant wrote: Wed Jul 22, 2020 9:47 am Michael Kitces has a new blog post that does a great job illustrating issues related to Roth conversions and harvesting capital gains. Here's a link:

https://www.kitces.com/blog/navigating- ... nversions/

I especially appreciate Kitces drawing attention to the reality that any opportunity to do Roth conversions in the 12% bracket is matched to an opportunity to realize capital gains in the 0% LTCG bracket. He explains thoroughly how low income tax brackets and the 0% capital gains bracket have a shared space that can only be used for one or the other or a finite combination of both. He shows that, at the very least, investors should realize ordinary income to fill the standard deduction or other deductions first, then consider capital gains vs. ordinary income (like Roth conversions) in higher brackets.

Unfortunately, Kitces appears to assume that the default priority should be to realize capital gains tax at the 0% rate rather than recognize ordinary income tax at the 10 and 12% rates. Kitces appears focused on the fact that capital gains taxed at 0% have a 15% differential from the next bracket, while ordinary income only has a 10% differential between the 12% and 22% brackets. While this may be true for some investors, it is problematic for four reasons.

One, investors can choose to defer capital gains as long as they wish, but they can't defer IRA income forever. Investors thus have much more control over capital gains once they begin taking social security, but RMDs can cause taxability in social security regardless of investor wishes. Realizing as much ordinary income as possible before social security begins can therefore reduce social security taxation in the majority of years while still leaving the investor in control of when to realize capital gains. While I haven't quantified it--and it would depend on the investor--I believe this means the savings from realizing ordinary income through a Roth conversion before SS is more than the 10% bracket differential between 12% and 22%. Kitces does not mention this directly.

Two, if investors do need additional money beyond retirement account withdrawals, social security, and other sources in the future, capital gains even taxed at the 15% rate are more efficient since any realization of capital gains comes with a return of basis. In other words, to get $100 out of the IRA, the investor has to pay taxes on the whole amount. To get $100 out of a taxable account, the investor will only pay taxes on the gains. Further, the investor can target the most desirable ratio based on tax lots available in the account. Capital gains assets are more efficient for obtaining a specific amount of cash with minimum taxation while affording significant flexibility. Kitces does not mention this; in fact, his example assumes a zero basis.

Three, and this is the most important one, assets with capital gains enjoy the step-up basis at death, while IRA balances do not. An investor really will be wasting time recognizing gains at 0% if the investor could have just realized ordinary income from the IRA while allowing the step-up basis to eliminate capital gains taxes when leaving assets to heirs. Fortunately, Kitces lists this as a "mitigating factor," but I think it deserves to be at the top of the considerations, not buried at the bottom.

Fourth, recognizing ordinary income through a Roth conversion allows all future growth to be free of taxes, while recognizing capital gains does not. In other words, if I perform a Roth conversion to move $100 from the IRA to the Roth IRA, I no longer have any tax liability from future growth in the asset. If it goes up 7%, I get the full 7%. For those who like the "government loan" analogy for IRAs, I have fully and finally paid off the loan. Not so for realizing capital gains. That just pays the tax bill for a specific nominal dollar amount, but it doesn't relieve the investor of taxes on additional growth in the future. While the relationship between this issue and risk can be complex, I nevertheless think getting the money in the Roth account can create a net present value of savings from avoiding taxes on future growth compared to the taxable account.

Overall, Kitces does a great job explaining these issues in the article. I just think he makes a naive assumption that the 15% differential between capital gains taxes trumps the 10% differential for ordinary income. The default, in my opinion, really should be to recognize ordinary income up to the top of the 12% bracket in years before SS starts, and that should only really shift for more specific situations. Beyond the 12% bracket, obviously, it gets more complex as the bracket differentials and IRMAA issues require more specific planning.
Nice analysis.

Do you think any of that changes (or if either is even worthwhile) if one is an early retiree withdrawing in 12% bracket, but also taking advantage of ACA Tax credits?
This raises the effective rates of either by 9.8%. I want to be doing Roth Conversions, but am still trying to convince myself it makes sense.
The retiree should really model this based on their specific circumstances, and I'll explain why. The structure of the subsidy is that a taxpayer is guaranteed to spend no more than a certain percentage of their income on a silver plan in their state, with the max percent varying as income increases. However, the subsidy completely disappears for income over the ACA "cliff." That means the marginal tax rate on income realized in the range from the minimum ACA income to the maximum ACA income or "cliff" is higher than the published tax rates. There my be a large "tax cost" at the cliff, but the marginal tax rate is back to normal above that. The marginal tax rates apply to both capital gains and ordinary income.

Thus, for example, the 2019 poverty guideline used for 2020 ACA insurance plans for a single individual was $12,490. The ACA minimum is thus $12,490 and the maximum is 400% of that or $49960. Say the benchmark silver plan for the year for the individual is $8,000. Somebody at the low end just above the minimum--let's say $14,000--is guaranteed to spend no more than 2% of income, so they would get a subsidy of 8000-14000*.02 = 7720. If that person has more income to reach $20,000, they will be under the 5% guarantee. Then, the subsidy is 8000-20000*.05 = 7000. Further, that whole change was in the 10% bracket. So realizing more income cost (20000-14000)*.10+(7720-7000) = 1320, or 1320/6000 = 22%. A higher marginal tax on that range indeed. Going up the chain, let's say the individual makes $49959, just before the cliff. That person receives a subsidy of 8000-49959*.095 = 3253.90. If they make a couple more dollars, they lose this whole amount. Generally, this is only going to be worth it if the taxpayer has a rather large tax benefit. For example, if they're avoiding future 32% income by placing it in the 22% bracket, they would be realizing 10% in tax arbitrage on everything over ~$52,000, but they would need to recognize that on over $32000 just to break even! So you're right, it's a big deterrence. As a rule, then, we can say that the taxpayer needs to realize income (say from Roth conversions) a bit past the standard deduction just to qualify for ACA subsidy premiums, but it's more complicated past that threshold.

Here are three things, though, about why the taxpayer needs to model it in specific circumstances. First, the benchmark determining the subsidy can make a large difference. The marginal tax rate impact itself has nothing to do with the size of the subsidy until you hit the cliff because the value of the subsidy is the benchmark cost of insurance less the maximum share of income, which doesn't vary based on the benchmark. The benchmark cost of insurance could vary widely from state to state. Thus, every ACA subsidy recipient faces the same cost until they hit the cliff OR their particular state benchmark is has been exhausted. This can change the impact dramatically. In the example above, we used $8000. That made sure the built-in tax rates were added and a big pain was felt at the cliff. However, imagine that the benchmark was $4000. In that case, the subsidy would completely phase out before the cliff. If b-i*m=s, where b is benchmark, i is income, m is maximum percent, and s is subsidy, we can solve for i (postulating that it's already in the top 9.5% max zone) as follows: 4000-.095i=0, .095i=4000, i=4000/.095, i=42105. So in that state, the retiree loses the subsidy completely before ever hitting the cliff, and they have 12% space left to boot. In that case, it is more likely that the taxpayer should eat the slightly higher tax rates and fill out the 12%. If the benchmark was $2000, the subsidy is practically just a speedbump. So generally, then, the state benchmark can have big consequences for the value of conversions.

Second, notice that realizing income after age 70 can be more destructive than realizing it in the face of ACA subsidy loss, depending on the size of the benchmark plans and the size of social security. After age 70, realizing income can trigger SS taxation up to 85% of taxable social security benefits. If social security for an individual taxpayer would be $30,000 per year, income taxes in the 12% bracket for 85% of this amount could be $3060. The retiree is more likely to face this tax if they wait on Roth conversions and take RMDs instead. So if the question is whether to lose a $3060 ACA subsidy for five years from age 60 to 64 to do Roth conversions or be forced to take RMDs starting at age 72 that cause all of social security to be taxed for $3060, well, it's complicated. In other words, both SS and ACA subsidies create an incentive to push other income sources to other years to avoid taxes--once they're both pushing income sources out, you might have to pick between them, or see what you can do to push all of the income to other years (like age 65-69, with the caveat of IRMAA).

Third, note that the tax brackets and ACA subsidies vary materially by household size. In the example above, the ACA minimum and ACA maximum closely match the bottom of the 10% bracket and the top of the 12% bracket for a single individual. That meant that the ACA cliff was close to the 22% bracket, so it's harder to justify going over the cliff. It's different for a MFJ couple. For them, the ACA minimum is below the bottom of the 10% bracket near $17,000, but the max is squarely inside the 12% bracket at $67640. So, the MFJ couple would probably see a more severe ACA cliff since the maximum cost went from .095*49960=4760 to .095*67640=6426, but the subsidy approximately doubled because the number of persons covered doubled. On the other hand, the MFJ couple has almost $40000 of 12% bracket space left for realizing income or capital gains. So in the states with the lowest benchmarks, MFJ couples are going to be much more likely to want to blow past the ACA cliff, while in states with the highest benchmarks, they're going to be more deterred.

Taking these into account, yeah, ACA subsidies can deter realizing income that you would otherwise want to realize. But whether to do so anyway depends heavily on the number of years left, household size, the state benchmark, and various factors affecting SS claiming like total wealth, spousal ages, etc. This is an area that is ripe for a very sophisticated calculator like Open Social Security or i-ORP to optimize (or for a good financial planner to make a detailed spreadsheet).
Last edited by petulant on Wed Jul 22, 2020 5:37 pm, edited 1 time in total.
Topic Author
petulant
Posts: 1901
Joined: Thu Sep 22, 2016 1:09 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by petulant »

FIREchief wrote: Wed Jul 22, 2020 3:06 pm That's a good article. It's rare (and difficult) that somebody tries to cover the broader picture including 0% LTCG tax gain harvesting, Roth conversions, IRMAA, etc. in a single article. That said, he only has one example of MFJ, and it doesn't get into the much more difficult "what if" planning for one of the spouses dying early, where the surviving spouse's marginal rate may spike but they may pay no more in LTCG taxes.

It's timely for me because I've recently been running some comparisons to attempt to prioritize tax gain harvesting vs. Roth conversions. Unfortunately, we really have no idea if we'll really revert to the pre-2018 federal tax structure after 2025 (as is currently the law). I see the variables out of Washington as harder to deal with than market return forecasts or life expectancies (which are all of course unknowable). I can plan as if DW and I will both live to 99 and require collectively ten years of LTC. I can also adopt a conservative market return forecast. Both of those will add "safety" to any plans. I can't really do something similar with the tax code in the future.
What's material if the tax code reverts in 2025? All I'm thinking of is switch the 12% to 15% and 22% to 25% etc., but it seems like most of the decisions made under either set of rates still mostly applies except in some edge cases.
Topic Author
petulant
Posts: 1901
Joined: Thu Sep 22, 2016 1:09 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by petulant »

shess wrote: Wed Jul 22, 2020 5:07 pm
petulant wrote: Wed Jul 22, 2020 9:47 am Michael Kitces has a new blog post that does a great job illustrating issues related to Roth conversions and harvesting capital gains. Here's a link:

https://www.kitces.com/blog/navigating- ... nversions/

I especially appreciate Kitces drawing attention to the reality that any opportunity to do Roth conversions in the 12% bracket is matched to an opportunity to realize capital gains in the 0% LTCG bracket. He explains thoroughly how low income tax brackets and the 0% capital gains bracket have a shared space that can only be used for one or the other or a finite combination of both. He shows that, at the very least, investors should realize ordinary income to fill the standard deduction or other deductions first, then consider capital gains vs. ordinary income (like Roth conversions) in higher brackets.
Thanks for the link!

Last year, I did this analysis for us. I'm 51, my spouse is 49, so we have plenty of time before RMDs. We have some substantial positions with significant LTCG, so initially I was strongly tilted towards harvesting those LTCG. BUT, after working things back and forth, it really felt like Roth conversions were worth more over the entire lifecycle, especially if one of us were to die earlier, forcing distributions to accelerate.

But this year's shenanigans demonstrated to me that I was pushing my luck on our asset allocation. I was making an argument against rebalancing, because I was greedy (basically). So I've made some sales to take advantage of the market and get things in order, and now I'm thinking I should probably run with that to some extent. Still working the numbers, but it feels like swapping some of our taxable holdings from equities to fixed-income, and vice versa in tax-deferred, makes sense insofar as that taxable fixed-income is then drawn down over the next few years to support living expenses. That in turn lets us increase Roth conversions over the next few years, because we won't have to balance against LTCG sales.

I'm almost tempted to build a tax-estimator module so that I can just write programs to run scenarios and generate graphs and heatmaps. It feels like the spreadsheet-based systems end up having self-referential issues, at least in Google Sheets. It would be interesting if a constraint on the tax code would be a requirement for the IRS to publish a code module to implement things :-).
That's an entirely reasonable thing to tilt the scales. If your taxable account is causing you to have too much stock for your comfort, it certainly makes sense to harvest gains there. Running calculators is best, but I have to suggest placing all traditional IRA balances in bonds first so that it doesn't grow, then see how much reallocating to do in taxable while leaving Roth in stocks.
Topic Author
petulant
Posts: 1901
Joined: Thu Sep 22, 2016 1:09 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by petulant »

deltaneutral83 wrote: Wed Jul 22, 2020 3:37 pm So on the surface mathematically, the spread between LTCG at 0 and 15% is greater and gives the appearance of being the better option (for those retired and not yet on SS) against Roth conversions up to the 12% bucket (because the next bucket is 22% creating only 10% spread) but in the end Roth conversions up to the 12% bucket (assuming MFJ) is more than likely to be the better option over 0% LTCG for a host of other reasons?
What I am arguing specifically is that if you were giving people a default, Roth conversions to the top of the 12% should be the default for the reasons mentioned. However, it is very clear that a specific analysis is the best approach.
User avatar
FiveK
Posts: 10290
Joined: Sun Mar 16, 2014 2:43 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FiveK »

petulant wrote: Wed Jul 22, 2020 5:23 pm Taking these into account, yeah, ACA subsidies can deter realizing income that you would otherwise want to realize. But whether to do so anyway depends heavily on the number of years left, household size, the state benchmark, and various factors affecting SS claiming like total wealth, spousal ages, etc. This is an area that is ripe for a very sophisticated calculator like Open Social Security or i-ORP to optimize (or for a good financial planner to make a detailed spreadsheet).
Very sophisticated indeed. :)

It really is a numerically difficult problem to solve due to all the real world non-linearites, and AFAIK there is no tool commercially available that considers all the common choices and inputs.

For long term strategies, one might use Open Social Security, Extended I-ORP and the Retiree Portfolio Model.
For "this year" tactics, the marginal rate chart available in the personal finance toolbox Excel spreadsheet can be useful.
shess
Posts: 663
Joined: Wed May 17, 2017 12:02 am

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by shess »

petulant wrote: Wed Jul 22, 2020 5:29 pm That's an entirely reasonable thing to tilt the scales. If your taxable account is causing you to have too much stock for your comfort, it certainly makes sense to harvest gains there. Running calculators is best, but I have to suggest placing all traditional IRA balances in bonds first so that it doesn't grow, then see how much reallocating to do in taxable while leaving Roth in stocks.
The traditional IRA funds are mostly in international stocks, which seem to be just as good as bond funds for keeping growth under control. But, yeah, shifting where assets live does mean revisiting the overall layout to make sure that the items expected to generate more growth are in the right places.

I've been pretty happy with what i-ORP tells me on this front, except that our taxable account contains a position of employer stock with is substantial and well over 99% LTCG. So the default outputs require some interpretation, because I think there's a good likelihood we won't have to touch that position and can leave it to charity or to heirs with step-up. Mostly I try running things with different input scenarios to get a sense of where different options make sense - but it would sure be cool to be able to just see the dimensions and determine which ones to hold fixed at any given run.
User avatar
FIREchief
Posts: 5309
Joined: Fri Aug 19, 2016 6:40 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FIREchief »

kaneohe wrote: Wed Jul 22, 2020 4:33 pm
FIREchief wrote: Wed Jul 22, 2020 4:13 pm ........................................................................................

The numbers change a bit as DW and I reach age 63 and IRMAA enters the equation. I think the article suggested that IRMAA "only" adds a 1% equivalent tax surcharge, but the current numbers are 1.58%, 1.93% and 1.93% for the first three IRMAA tiers. Not prohibitive, but worth consideration.
Can you show how that is calculated...........I seem to be getting more like 5%........
My bad. I corrected my post. I forgot to double for DW and I and I also neglected to include Part D. :oops: :oops:
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.
User avatar
FIREchief
Posts: 5309
Joined: Fri Aug 19, 2016 6:40 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FIREchief »

(deleted)
Last edited by FIREchief on Wed Jul 22, 2020 7:31 pm, edited 1 time in total.
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.
User avatar
FIREchief
Posts: 5309
Joined: Fri Aug 19, 2016 6:40 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FIREchief »

petulant wrote: Wed Jul 22, 2020 5:23 pm Thus, for example, the 2019 poverty guideline used for 2020 ACA insurance plans for a single individual was $12,490. The ACA minimum is thus $12,490
Isn't it 139% of the federal poverty level in states that have expanded Medicaid?
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.
User avatar
FIREchief
Posts: 5309
Joined: Fri Aug 19, 2016 6:40 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FIREchief »

petulant wrote: Wed Jul 22, 2020 5:24 pm
FIREchief wrote: Wed Jul 22, 2020 3:06 pm That's a good article. It's rare (and difficult) that somebody tries to cover the broader picture including 0% LTCG tax gain harvesting, Roth conversions, IRMAA, etc. in a single article. That said, he only has one example of MFJ, and it doesn't get into the much more difficult "what if" planning for one of the spouses dying early, where the surviving spouse's marginal rate may spike but they may pay no more in LTCG taxes.

It's timely for me because I've recently been running some comparisons to attempt to prioritize tax gain harvesting vs. Roth conversions. Unfortunately, we really have no idea if we'll really revert to the pre-2018 federal tax structure after 2025 (as is currently the law). I see the variables out of Washington as harder to deal with than market return forecasts or life expectancies (which are all of course unknowable). I can plan as if DW and I will both live to 99 and require collectively ten years of LTC. I can also adopt a conservative market return forecast. Both of those will add "safety" to any plans. I can't really do something similar with the tax code in the future.
What's material if the tax code reverts in 2025? All I'm thinking of is switch the 12% to 15% and 22% to 25% etc., but it seems like most of the decisions made under either set of rates still mostly applies except in some edge cases.
There are other possibilities. The facts are that most of the 2018 changes are set to expire after 2025. Sure, current law leads to what you suggest. The challenge is, that this is only one possible outcome. I think we need to acknowledge that uncertainty, regardless of how we wish or predict that it will play out. I'm pretty confident that the Roth laws will remain stable (famous last words :oops: ) but not as confident in several other aspects (LTCG treatment, etc.).
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.
User avatar
dodecahedron
Posts: 5355
Joined: Tue Nov 12, 2013 12:28 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by dodecahedron »

I will not be harvesting any capital gains. I have accumulated capital losses (from past TLH) which I plan to slowly draw down as an annual $3K per year offset to ordinary income.

I do not care to waste of any of my accumulate TLH against realized capital gains.
User avatar
FiveK
Posts: 10290
Joined: Sun Mar 16, 2014 2:43 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FiveK »

FIREchief wrote: Wed Jul 22, 2020 6:11 pm
FiveK wrote: Wed Jul 22, 2020 4:52 pm
kaneohe wrote: Wed Jul 22, 2020 4:33 pm Can you show how that is calculated...........I seem to be getting more like 5%........
+1

At best it would be 3.8%, 4.7%, and 4.7% to go over the $87K, $109K, and $136K tiers respectively for Medicare B and D - and that's if one goes all the way up exactly to the next tier.

E.g.,
12*((202.4+12.2)-(144.6+0.00))/(109000-87000) = 3.8%
12*((289.2+31.5)-(202.4+12.2))/(136000-109000) = 4.7%
If I'm reading the Medicare website correctly, the part D charges are all deltas from your monthly Part B, so I don't believe you should have subtracted that $12.20 in your second example. Of course, my original numbers were way further from correct. :P
Based on Medicare costs at a glance | Medicare, it seems the income-dependent monthly costs for a single filer are

Code: Select all

       Part B  Part D  Total
Base    144.6   0      144.6
Tier 1  202.4   12.2   214.6
Tier 2  289.2   31.5   320.7
Tier 3  376.0   50.7   426.7
...
The cost of going to a higher tier would then be the difference between the totals of the adjoining tiers. Or am I missing something?
User avatar
FIREchief
Posts: 5309
Joined: Fri Aug 19, 2016 6:40 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FIREchief »

(deleted)
Last edited by FIREchief on Wed Jul 22, 2020 7:31 pm, edited 1 time in total.
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.
User avatar
FiveK
Posts: 10290
Joined: Sun Mar 16, 2014 2:43 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FiveK »

FIREchief wrote: Wed Jul 22, 2020 6:48 pm
FiveK wrote: Wed Jul 22, 2020 6:40 pm Based on Medicare costs at a glance | Medicare, it seems the income-dependent monthly costs for a single filer are

Code: Select all

       Part B  Part D  Total
Base    144.6   0      144.6
Tier 1  202.4   12.2   214.6
Tier 2  289.2   31.5   320.7
Tier 3  376.0   50.7   426.7
...
The cost of going to a higher tier would then be the difference between the totals of the adjoining tiers. Or am I missing something?
I think you are, but I'm not certain. If you expand the detailed cost information for 2020 for Part D is reads:
Base:
your plan premium
Tier 2:
$31.50 + your plan premium
I read it to indicate that I'll pay $31.50 more for Part D if my AGI falls within the range that my Part B costs $289.20.
Agreed on "pay $31.50 more for Part D if my AGI falls within the range that my Part B costs $289.20."

What about "pay $12.20 more for Part D if my AGI falls within the range that my Part B costs $202.40?"
User avatar
FIREchief
Posts: 5309
Joined: Fri Aug 19, 2016 6:40 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FIREchief »

(deleted)
Last edited by FIREchief on Wed Jul 22, 2020 7:31 pm, edited 1 time in total.
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.
User avatar
FiveK
Posts: 10290
Joined: Sun Mar 16, 2014 2:43 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FiveK »

FIREchief wrote: Wed Jul 22, 2020 7:08 pm
FiveK wrote: Wed Jul 22, 2020 6:51 pm
FIREchief wrote: Wed Jul 22, 2020 6:48 pm
FiveK wrote: Wed Jul 22, 2020 6:40 pm Based on Medicare costs at a glance | Medicare, it seems the income-dependent monthly costs for a single filer are

Code: Select all

       Part B  Part D  Total
Base    144.6   0      144.6
Tier 1  202.4   12.2   214.6
Tier 2  289.2   31.5   320.7
Tier 3  376.0   50.7   426.7
...
The cost of going to a higher tier would then be the difference between the totals of the adjoining tiers. Or am I missing something?
I think you are, but I'm not certain. If you expand the detailed cost information for 2020 for Part D is reads:
Base:
your plan premium
Tier 2:
$31.50 + your plan premium
I read it to indicate that I'll pay $31.50 more for Part D if my AGI falls within the range that my Part B costs $289.20.
Agreed on "pay $31.50 more for Part D if my AGI falls within the range that my Part B costs $289.20."

What about "pay $12.20 more for Part D if my AGI falls within the range that my Part B costs $202.40?"
Sure. That's how I read it. It's $x.xx more than your plan premium where "your plan premium" is what you pay for a base plan without IRMAA.
OK. Then to go from Tier 1 to Tier 2 it costs an extra (289.2+31.5)-(202.4+12.2) = 320.7 - 214.6 = $106.1/mo, correct?
User avatar
FIREchief
Posts: 5309
Joined: Fri Aug 19, 2016 6:40 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FIREchief »

FiveK wrote: Wed Jul 22, 2020 7:17 pm
OK. Then to go from Tier 1 to Tier 2 it costs an extra (289.2+31.5)-(202.4+12.2) = 320.7 - 214.6 = $106.1/mo, correct?
You're right. My brain is just not having a good day. :oops:
Last edited by FIREchief on Wed Jul 22, 2020 7:32 pm, edited 1 time in total.
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.
User avatar
FiveK
Posts: 10290
Joined: Sun Mar 16, 2014 2:43 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FiveK »

I shan't be casting any stones. ;) :sharebeer
Topic Author
petulant
Posts: 1901
Joined: Thu Sep 22, 2016 1:09 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by petulant »

FIREchief wrote: Wed Jul 22, 2020 6:28 pm
petulant wrote: Wed Jul 22, 2020 5:23 pm Thus, for example, the 2019 poverty guideline used for 2020 ACA insurance plans for a single individual was $12,490. The ACA minimum is thus $12,490
Isn't it 139% of the federal poverty level in states that have expanded Medicaid?
Yes, sort of. The rule appears to be that they're eligible at 100%, but Medicaid eligibility disqualifies them regardless of income. Expanded Medicaid is separately set at 139% of FPL.
marcopolo
Posts: 3468
Joined: Sat Dec 03, 2016 10:22 am

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by marcopolo »

petulant wrote: Wed Jul 22, 2020 5:23 pm
marcopolo wrote: Wed Jul 22, 2020 1:47 pm
petulant wrote: Wed Jul 22, 2020 9:47 am Michael Kitces has a new blog post that does a great job illustrating issues related to Roth conversions and harvesting capital gains. Here's a link:

https://www.kitces.com/blog/navigating- ... nversions/

I especially appreciate Kitces drawing attention to the reality that any opportunity to do Roth conversions in the 12% bracket is matched to an opportunity to realize capital gains in the 0% LTCG bracket. He explains thoroughly how low income tax brackets and the 0% capital gains bracket have a shared space that can only be used for one or the other or a finite combination of both. He shows that, at the very least, investors should realize ordinary income to fill the standard deduction or other deductions first, then consider capital gains vs. ordinary income (like Roth conversions) in higher brackets.

Unfortunately, Kitces appears to assume that the default priority should be to realize capital gains tax at the 0% rate rather than recognize ordinary income tax at the 10 and 12% rates. Kitces appears focused on the fact that capital gains taxed at 0% have a 15% differential from the next bracket, while ordinary income only has a 10% differential between the 12% and 22% brackets. While this may be true for some investors, it is problematic for four reasons.

One, investors can choose to defer capital gains as long as they wish, but they can't defer IRA income forever. Investors thus have much more control over capital gains once they begin taking social security, but RMDs can cause taxability in social security regardless of investor wishes. Realizing as much ordinary income as possible before social security begins can therefore reduce social security taxation in the majority of years while still leaving the investor in control of when to realize capital gains. While I haven't quantified it--and it would depend on the investor--I believe this means the savings from realizing ordinary income through a Roth conversion before SS is more than the 10% bracket differential between 12% and 22%. Kitces does not mention this directly.

Two, if investors do need additional money beyond retirement account withdrawals, social security, and other sources in the future, capital gains even taxed at the 15% rate are more efficient since any realization of capital gains comes with a return of basis. In other words, to get $100 out of the IRA, the investor has to pay taxes on the whole amount. To get $100 out of a taxable account, the investor will only pay taxes on the gains. Further, the investor can target the most desirable ratio based on tax lots available in the account. Capital gains assets are more efficient for obtaining a specific amount of cash with minimum taxation while affording significant flexibility. Kitces does not mention this; in fact, his example assumes a zero basis.

Three, and this is the most important one, assets with capital gains enjoy the step-up basis at death, while IRA balances do not. An investor really will be wasting time recognizing gains at 0% if the investor could have just realized ordinary income from the IRA while allowing the step-up basis to eliminate capital gains taxes when leaving assets to heirs. Fortunately, Kitces lists this as a "mitigating factor," but I think it deserves to be at the top of the considerations, not buried at the bottom.

Fourth, recognizing ordinary income through a Roth conversion allows all future growth to be free of taxes, while recognizing capital gains does not. In other words, if I perform a Roth conversion to move $100 from the IRA to the Roth IRA, I no longer have any tax liability from future growth in the asset. If it goes up 7%, I get the full 7%. For those who like the "government loan" analogy for IRAs, I have fully and finally paid off the loan. Not so for realizing capital gains. That just pays the tax bill for a specific nominal dollar amount, but it doesn't relieve the investor of taxes on additional growth in the future. While the relationship between this issue and risk can be complex, I nevertheless think getting the money in the Roth account can create a net present value of savings from avoiding taxes on future growth compared to the taxable account.

Overall, Kitces does a great job explaining these issues in the article. I just think he makes a naive assumption that the 15% differential between capital gains taxes trumps the 10% differential for ordinary income. The default, in my opinion, really should be to recognize ordinary income up to the top of the 12% bracket in years before SS starts, and that should only really shift for more specific situations. Beyond the 12% bracket, obviously, it gets more complex as the bracket differentials and IRMAA issues require more specific planning.
Nice analysis.

Do you think any of that changes (or if either is even worthwhile) if one is an early retiree withdrawing in 12% bracket, but also taking advantage of ACA Tax credits?
This raises the effective rates of either by 9.8%. I want to be doing Roth Conversions, but am still trying to convince myself it makes sense.
The retiree should really model this based on their specific circumstances, and I'll explain why. The structure of the subsidy is that a taxpayer is guaranteed to spend no more than a certain percentage of their income on a silver plan in their state, with the max percent varying as income increases. However, the subsidy completely disappears for income over the ACA "cliff." That means the marginal tax rate on income realized in the range from the minimum ACA income to the maximum ACA income or "cliff" is higher than the published tax rates. There my be a large "tax cost" at the cliff, but the marginal tax rate is back to normal above that. The marginal tax rates apply to both capital gains and ordinary income.

Thus, for example, the 2019 poverty guideline used for 2020 ACA insurance plans for a single individual was $12,490. The ACA minimum is thus $12,490 and the maximum is 400% of that or $49960. Say the benchmark silver plan for the year for the individual is $8,000. Somebody at the low end just above the minimum--let's say $14,000--is guaranteed to spend no more than 2% of income, so they would get a subsidy of 8000-14000*.02 = 7720. If that person has more income to reach $20,000, they will be under the 5% guarantee. Then, the subsidy is 8000-20000*.05 = 7000. Further, that whole change was in the 10% bracket. So realizing more income cost (20000-14000)*.10+(7720-7000) = 1320, or 1320/6000 = 22%. A higher marginal tax on that range indeed. Going up the chain, let's say the individual makes $49959, just before the cliff. That person receives a subsidy of 8000-49959*.095 = 3253.90. If they make a couple more dollars, they lose this whole amount. Generally, this is only going to be worth it if the taxpayer has a rather large tax benefit. For example, if they're avoiding future 32% income by placing it in the 22% bracket, they would be realizing 10% in tax arbitrage on everything over ~$52,000, but they would need to recognize that on over $32000 just to break even! So you're right, it's a big deterrence. As a rule, then, we can say that the taxpayer needs to realize income (say from Roth conversions) a bit past the standard deduction just to qualify for ACA subsidy premiums, but it's more complicated past that threshold.

Here are three things, though, about why the taxpayer needs to model it in specific circumstances. First, the benchmark determining the subsidy can make a large difference. The marginal tax rate impact itself has nothing to do with the size of the subsidy until you hit the cliff because the value of the subsidy is the benchmark cost of insurance less the maximum share of income, which doesn't vary based on the benchmark. The benchmark cost of insurance could vary widely from state to state. Thus, every ACA subsidy recipient faces the same cost until they hit the cliff OR their particular state benchmark is has been exhausted. This can change the impact dramatically. In the example above, we used $8000. That made sure the built-in tax rates were added and a big pain was felt at the cliff. However, imagine that the benchmark was $4000. In that case, the subsidy would completely phase out before the cliff. If b-i*m=s, where b is benchmark, i is income, m is maximum percent, and s is subsidy, we can solve for i (postulating that it's already in the top 9.5% max zone) as follows: 4000-.095i=0, .095i=4000, i=4000/.095, i=42105. So in that state, the retiree loses the subsidy completely before ever hitting the cliff, and they have 12% space left to boot. In that case, it is more likely that the taxpayer should eat the slightly higher tax rates and fill out the 12%. If the benchmark was $2000, the subsidy is practically just a speedbump. So generally, then, the state benchmark can have big consequences for the value of conversions.

Second, notice that realizing income after age 70 can be more destructive than realizing it in the face of ACA subsidy loss, depending on the size of the benchmark plans and the size of social security. After age 70, realizing income can trigger SS taxation up to 85% of taxable social security benefits. If social security for an individual taxpayer would be $30,000 per year, income taxes in the 12% bracket for 85% of this amount could be $3060. The retiree is more likely to face this tax if they wait on Roth conversions and take RMDs instead. So if the question is whether to lose a $3060 ACA subsidy for five years from age 60 to 64 to do Roth conversions or be forced to take RMDs starting at age 72 that cause all of social security to be taxed for $3060, well, it's complicated. In other words, both SS and ACA subsidies create an incentive to push other income sources to other years to avoid taxes--once they're both pushing income sources out, you might have to pick between them, or see what you can do to push all of the income to other years (like age 65-69, with the caveat of IRMAA).

Third, note that the tax brackets and ACA subsidies vary materially by household size. In the example above, the ACA minimum and ACA maximum closely match the bottom of the 10% bracket and the top of the 12% bracket for a single individual. That meant that the ACA cliff was close to the 22% bracket, so it's harder to justify going over the cliff. It's different for a MFJ couple. For them, the ACA minimum is below the bottom of the 10% bracket near $17,000, but the max is squarely inside the 12% bracket at $67640. So, the MFJ couple would probably see a more severe ACA cliff since the maximum cost went from .095*49960=4760 to .095*67640=6426, but the subsidy approximately doubled because the number of persons covered doubled. On the other hand, the MFJ couple has almost $40000 of 12% bracket space left for realizing income or capital gains. So in the states with the lowest benchmarks, MFJ couples are going to be much more likely to want to blow past the ACA cliff, while in states with the highest benchmarks, they're going to be more deterred.

Taking these into account, yeah, ACA subsidies can deter realizing income that you would otherwise want to realize. But whether to do so anyway depends heavily on the number of years left, household size, the state benchmark, and various factors affecting SS claiming like total wealth, spousal ages, etc. This is an area that is ripe for a very sophisticated calculator like Open Social Security or i-ORP to optimize (or for a good financial planner to make a detailed spreadsheet).

Thanks for the detailed answers. While it is very complicated, I am glad to see I am thinking about it the right way. In our circumstances (state specific) the ACA tax credits end up being quite large (~20K/year), with our MAGI being close to the top of 25% bracket (our state also has higher FPL). And our cliff is steep. All my analysis shows a slight advantage to NOT doing any Roth conversion or TGH at this time. So, we are just withdrawing living expenses from taxable account, with no other tax moves.

I suspect at some point the tax credits will either be eliminated or curtailed based on assets as well as income. We will revisit doing Roth Conversions at that time. I do agree with you that given the choice Roth Conversions are more valuable than TGH.

Thanks again for your insight.
Once in a while you get shown the light, in the strangest of places if you look at it right.
Topic Author
petulant
Posts: 1901
Joined: Thu Sep 22, 2016 1:09 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by petulant »

This morning I wanted to come back to the fourth advantage I listed, which is that a Roth conversion frees the tax-deferred balance from tax on future growth while realizing capital gains does not. So, even though realizing the capital gains may save a higher differential in the short run (15% rather than 22%-12%=10%), it doesn't account for the taxation of future growth. I wanted to put some numbers to that insight.

Start with a basic situation, taking out all other issues. I have $100 invested in stocks in a tax-deferred account and $100 in unrealized capital gains in a stock position. Let's say I could realize the income from a Roth conversion and pay 12%, or I could realize the capital gains and pay 0%. After I choose, both accounts will remain invested in stocks, growing at 7% per year. As a baseline, if I do nothing and let my 12% space go to waste this year, the $100 in IRA stocks will grow at 7% to $197 in 10 years, but it will be taxed at 22% at that point due to my assumption that the 12% space is finite--that is, I have other income/gains to fill all 12% space into the future. I will thus have a $43 liability and only $154 after tax. Further, my $100 in unrealized capital gains would grow at 6.8% after tax drag (assuming 2% dividend yield will be taxed at 0% LTCG rates, but crowding out other opportunities in the 12/0% bracket will cost at a minimum 10% in lost tax arbitrage, so .02*.10=.0020%) to $193, which after paying capital gains taxes in the future would be $167 [remember dividends created basis]. If I do nothing, I have $321 in cash in the future.

Let's say I harvest gains instead of doing nothing. If I leave the $100 in the IRA and have to pay 22% in the future, I will have $197 in 10 years with a ~$43 tax liability for a net ~$154. So that that point, not only did I pay an extra 10% on the $100 since I didn't take advantage of arbitrage, but I also paid the full 22% on all of the growth--almost another $100!

In the meantime, even though I realized the capital gains in my taxable account and left them invested, the future gains are also taxable. So I paid $0 to transform the gains into a new basis for my stock. But with the stock growing at 7% per year, the amount will be higher. Further, we need to reduce the growth rate by, say, .20% to account for dividend tax drag, so return of 6.8%. (Assuming 2% dividend yield will be taxed at 0% LTCG rates, but crowding out other opportunities in the 12/0% bracket will cost at a minimum 10% in lost tax arbitrage, so .02*.10=.0020%). The taxable stock position will thus grow to $193, of which $68 93 would be taxable at 15% (dividends over the years added $25 in basis). That's a net of 125+68*.85=$183 100+93*.85=$179.

So, assuming I had the option in an earlier year to engage in tax arbitrage using either a Roth conversion or realizing a capital gain, and then in the future I have other income to fill the 12% bracket, I have a total of $337 333 in funds I can access. That's better than doing nothing.

On the other hand, if I had performed a Roth conversion, it would have cost more up front and left me with $88 in the account. However, that amount would have grown at 7% without any future taxation. Thus, in my future year, I have a full $173 available. If I look at the capital gains, though, my $100 still grew to $193, but now more if it is taxable, leaving me with 25+168*.85=168 193*.85=$164. Overall, I have access to $341 354.

So, if 12%/0% space is actually scarce and I need to use it judiciously because marginal income will be realized above that amount in the future, I take it that Roth conversions are more effective.

Hold on, some might say. I noted in my earlier post that the relationship between these expectations and risk can be complex. Well, when I kept the stocks in my tax-deferred account, I had lower future returns and less risk than in the Roth account. My future gains were proportionately reduced by future taxation, while my future losses were mitigated by reduced taxation. On the other hand, if we assume parity of tax treatment between losses and gains (which isn't necessarily true, since loss claiming is limited but can also be claimed at a higher tax rate than gains), the risk-reward profile for the taxable stock position is roughly the same regardless of whether I realize the capital gains or not. So part of what I am doing in the Roth conversion is actually taking on more risk and a higher expected return.

For example, if I postulate that all future gains in the IRA would be taxed or lose money in the 22% bracket, the true expected return for stocks might be (1.07/1.02-1)*(1-.22)+.02=5.8%. (You should consider an expected inflation reduction here since future taxes on the IRA will be reduced by the widening of brackets for inflation. It's not just .07*(1-.22).) In the Roth account, the future expected returns might be a simple 7%. But both have proportionate risk to match their respective returns.

[While I originally engaged in the above exercise to equalize the expected return on stocks in a traditional IRA with the expected return on stocks in a Roth IRA, I mistakenly failed to consider that while a given dollar in the Roth is exposed to more risk, this exercise had fewer dollars actually placed in the Roth IRA. Thus, the relative risk of $88 in a Roth IRA in stocks and $100 in taxable stocks compared to $100 in a traditional IRA and $100 in taxable stocks is similar. I now believe that the original, "naive" comparison has more legitimacy.]

If we imagine for this illustration that bonds have real returns of zero and nominal returns that match tax bracket changes, we can identify a new asset allocation in the Roth that matches risk so we can adjudicate a bit more fairly. To reduce the return on $88 to 5.8%, we need .058/.07=.83 or 83% of our portfolio in stocks and the rest in bonds. So going back to the Roth conversion scenario, after converting in the 12% bracket, we have $88, of which we place 88*.83=$73 in stocks and $15 in bonds. This account will now grow to $158 in ten years. Added to $168 164, I have a total of $326. That's actually less than I had from realizing the capital gains early. The naive understanding that putting growth in the Roth account is better doesn't necessarily hold.

I'm not satisfied here, though. Having space in Roth seems like the most tax-advantaged way to hold stocks among all options. I don't like holding bonds in the Roth. So let's say, rather than using the precious Roth space to hold these bonds, I alter other holdings in my remaining tax-deferred space to bonds to equal the risk. I believe this is more akin to what a typical BH investor might do when juggling tax-deferred, Roth, and taxable accounts. In this example, we instead keep the Roth account entirely in stocks, exposing the portfolio to more risk than if we had realized capital gains and left the money in the traditional IRA. However, imagine that we then move $15/(1-.22)=$19 in the traditional account outside the $100 being converted from stocks to bonds, equalizing the risk. Here, the $19 in the traditional account would have grown to $29.50 after tax in ten years, while in bonds it would have become $18 after tax.

That means by performing a Roth conversion, leaving it in stocks, and equalizing risk by moving traditional space to bonds, we had $341+(18-29.50)=329.50 $354+(18-29.50)=$342.50. That does beat equalizing risk just inside the Roth account, but it still doesn't match harvesting gains. That does indeed beat both 1) realizing the capital gains tax up front AND 2) performing a Roth conversion, then equalizing risk within the Roth account.. Here are the outcomes collected again:

Roth conversion, all stocks more risk - $341 354 net cash
Realize capital gains - $337 333 net cash
Roth conversion, equalize risk in Trad'l IRA - $329.50 342.50 net cash
Roth conversion, equalize risk in Roth IRA - $326 322 net cash
Do nothing - $321 net cash

The power of looking at Roth conversions as creating an opportunity for a more efficient portfolio overall is powerful. Essentially what we are doing here is leveraging Roth space to more effectively manage future growth of tax liability in the traditional space. [Part of this depends on your opinion about what happens to risk after performing a Roth conversion and moving the stocks from the tax-deferred IRA to the Roth IRA. I now have the opinion that this does not increase risk when the tax liability on the conversion is paid out of an IRA withdrawal; in other words, a smaller amount of money in stocks in the Roth IRA directly offsets the higher risk in the Roth account, meaning the applicable comparison is between the "Roth conversion, all stocks" scenario and the "Realize capital gains" scenario.]

This can be very powerful when we recognize that stocks in a traditional account also create asymmetric risk/reward at some income levels where losses result in an inability to use all space in the lower bracket like 12%, while gains would be taxed at a higher bracket like 22%. Restricting exposure to this bracket near or in retirement can be very powerful.

Further, this benefit of Roth conversions over long-term capital gains depends on the relevant holding period. Essentially, the Roth conversion pays one upfront fee that makes the space more efficient for a long period. That's all the more reason to prioritize Roth conversions in *early* years.

Additionally, this consideration is intended to be combined with the other factors set out in the initial post, including an expectation of a step-up basis at death, the possibility of avoiding RMDs, and the flexibility of control over basis+capital gains in the future.

EDIT: Corrected a small point of math.
EDIT 2: Later on, I discovered an error where I did not take into account that the dividends in a taxable account, when reinvested, actually create new basis. I adjusted the calculations above and revised the outcome to take this into account. I also added a "do nothing" scenario. In the course of making these corrections, I determined my calculation of the Roth conversion strategy with no risk correction had an addition error. New material is underlined, while erroneous numbers have strikethroughs.
EDIT 3: I realized another error that offsets the error in EDIT 2. I performed a risk location mitigation that was unnecessary. Notes about this in the post are in italics and brackets.
Last edited by petulant on Tue Aug 11, 2020 5:34 pm, edited 7 times in total.
User avatar
House Blend
Posts: 4761
Joined: Fri May 04, 2007 1:02 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by House Blend »

Probably I am simply repeating arguments made in the article, but it is important to keep in mind that harvesting gains accomplishes nothing unless these are shares that will ultimately be spent during your lifetime. (Donations and bequests don't count as spending for this purpose.)

In any case, there's no point in gain-harvesting shares that will be passed to your children or grandchildren. Use that tax bracket space to Roth convert.
dodecahedron wrote: Wed Jul 22, 2020 6:39 pm I will not be harvesting any capital gains. I have accumulated capital losses (from past TLH) which I plan to slowly draw down as an annual $3K per year offset to ordinary income.

I do not care to waste of any of my accumulate TLH against realized capital gains.
+1.

I have substantial carryover losses. That means that with no gain harvesting and normal spending patterns, I have many years ahead of me in which I will pay 0% Fed and State on realized long and short cap gains, regardless of tax bracket. My attention will be focused on how much to Roth convert each year.
crefwatch
Posts: 655
Joined: Sun Apr 15, 2007 1:07 pm
Location: New Jersey, USA

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by crefwatch »

I found the article linked in the OP very useful. As I read about the topic on Bogleheads, I came across this article:

https://www.bogleheads.org/wiki/Traditi ... _cliffs.3F

It seems to me that it contributes to the fetishization of the IRMAA spikes. But if you look at the chart, the Cumulative tax line resumes a much lower number right after each IRMAA spike. That means, if you're paying $10,000 tax on a Roth conversion, the $1000 IRMAA cost is a minor aspect of it-assuming your brackets work our in the analysis.

It's also troubling to describe IRMAA as a "tax". It's much more of a dedicated, means-tested user fee for a very specific product every old person wants. It's a little like calling a bridge toll a "tax". If you hate taxes, you should love user fees. I pay IRMAA, by the way.
leftcoaster
Posts: 497
Joined: Mon Jul 23, 2007 4:04 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by leftcoaster »

What a great response to Kitces! A fine example of why this board is so valuable.

Another thing to consider is that rates of taxation are not permanently fixed. None of us has a crystal ball about the actions of future administrations, but we should be mindful that adjusting rates of taxation (and sun setting them) is a primary focus of our elected officials.
Topic Author
petulant
Posts: 1901
Joined: Thu Sep 22, 2016 1:09 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by petulant »

crefwatch wrote: Thu Jul 23, 2020 10:35 am I found the article linked in the OP very useful. As I read about the topic on Bogleheads, I came across this article:

https://www.bogleheads.org/wiki/Traditi ... _cliffs.3F

It seems to me that it contributes to the fetishization of the IRMAA spikes. But if you look at the chart, the Cumulative tax line resumes a much lower number right after each IRMAA spike. That means, if you're paying $10,000 tax on a Roth conversion, the $1000 IRMAA cost is a minor aspect of it-assuming your brackets work our in the analysis.

It's also troubling to describe IRMAA as a "tax". It's much more of a dedicated, means-tested user fee for a very specific product every old person wants. It's a little like calling a bridge toll a "tax". If you hate taxes, you should love user fees. I pay IRMAA, by the way.
There are several items like IRMAA, IBR student loan payments, and ACA subsidies that analytically function the same as marginal taxes, tax phase-outs, and tax cliffs, so a decision-maker should treat them all the same within their numerical limits. I don't think there's an intention to apply a political connotation doing so.
Last edited by petulant on Thu Jul 23, 2020 11:00 am, edited 1 time in total.
Hockey10
Posts: 838
Joined: Wed Aug 24, 2016 12:20 pm
Location: Philadelphia suburbs

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by Hockey10 »

OP, thanks for posting this as it is one of the better threads I have seen on Bogleheads.
User avatar
kramer
Posts: 1768
Joined: Wed Feb 21, 2007 2:28 am
Location: World Traveler

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by kramer »

petulant wrote: Wed Jul 22, 2020 9:47 am Fourth, recognizing ordinary income through a Roth conversion allows all future growth to be free of taxes, while recognizing capital gains does not. In other words, if I perform a Roth conversion to move $100 from the IRA to the Roth IRA, I no longer have any tax liability from future growth in the asset. If it goes up 7%, I get the full 7%. For those who like the "government loan" analogy for IRAs, I have fully and finally paid off the loan. Not so for realizing capital gains. That just pays the tax bill for a specific nominal dollar amount, but it doesn't relieve the investor of taxes on additional growth in the future. While the relationship between this issue and risk can be complex, I nevertheless think getting the money in the Roth account can create a net present value of savings from avoiding taxes on future growth compared to the taxable account.
As a single early retiree now in my 50's, this fourth factor was the slam dunk for me overwhelming everything else, getting future growth in the Roth, which I will probably not draw on for at least 20 more years.

As someone else mentioned, there are always more details to consider. I have more than the $300 in foreign taxes paid each year in my taxable account on Vanguard ETFs that track foreign indices, and via form 1116 I lose a fair portion of this (effectively forever) since the tax rate on my base investment income is not high enough to claim all of it. But when I do Roth conversions and increase my present tax bill, I get a higher portion of the foreign taxes paid back, and that lowers my conversion brackets from 10 and 12 to something like 8 and 9.5 So it makes it even an easier decision to do Roth conversions for now.

There is also something to be said for tax diversity ... when I retired I had almost nothing in a Roth, which has made Roth conversions a higher priority for me.
Angst
Posts: 2424
Joined: Sat Jun 09, 2007 11:31 am

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by Angst »

+1 Nice post. Thank you petulant.
HomeStretch
Posts: 4978
Joined: Thu Dec 27, 2018 3:06 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by HomeStretch »

+2

This thoughtful discussion is helpful to my financial modeling to determine the best way to maximize after-tax retirement income taking into account:
1. RMDs
2. possible tax gain harvesting
3. Utilizing tax loss carryforwards
4. possible Roth conversions
5. ACA subsidies
6. IRMAA
7. SS claiming strategy and SS taxation implications
8. Legacy bequests
9. Charitable donations
10. Gifts to children while alive
11. Tax rate changes sun setting after 2025

It can be a rabbit hole trying to take this all into account while avoiding model errors. I need to remind myself periodically that ‘good enough’ is okay and my model doesn’t need to be ‘perfect’.
User avatar
FIREchief
Posts: 5309
Joined: Fri Aug 19, 2016 6:40 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by FIREchief »

crefwatch wrote: Thu Jul 23, 2020 10:35 am It seems to me that it contributes to the fetishization of the IRMAA spikes. But if you look at the chart, the Cumulative tax line resumes a much lower number right after each IRMAA spike. That means, if you're paying $10,000 tax on a Roth conversion, the $1000 IRMAA cost is a minor aspect of it-assuming your brackets work our in the analysis.
Are you suggesting that we just ignore IRMAA thresholds in our financial planning?
It's also troubling to describe IRMAA as a "tax". It's much more of a dedicated, means-tested user fee for a very specific product every old person wants. It's a little like calling a bridge toll a "tax". If you hate taxes, you should love user fees. I pay IRMAA, by the way.
"means-tested-user-fee" sounds like a tax to me! :P
I am not a lawyer, accountant or financial advisor. Any advice or suggestions that I may provide shall be considered for entertainment purposes only.
MathWizard
Posts: 4329
Joined: Tue Jul 26, 2011 1:35 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by MathWizard »

02nz wrote: Wed Jul 22, 2020 10:01 am Good points. Related to this, I think in many experts' advice there's often a bias toward liquidating taxable accounts first and letting tax-advantaged grow (here's another example: https://www.morningstar.com/articles/84 ... ithdrawals), when in many cases it's actually better to first draw down/convert tax-deferred balances to manage taxes/RMDs, and leave the taxable for later in the hope of a step-up in basis (keeping in mind tax laws can and do change).
A middle ground that I plan to use is to do conversions to fill up the 12% bracket and using taxable to pay the taxes on the conversion
and to live on. I filled up ROTH and tax deferred first, and taxable only after maxing both. This means that taxable is not very large, won't
have lots of capital gains, so most of the draw from taxable will be free of tax.
Topic Author
petulant
Posts: 1901
Joined: Thu Sep 22, 2016 1:09 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by petulant »

livesoft wrote: Wed Jul 22, 2020 9:51 am Thanks for this. I think you should send your comment to Kitces or post it as a response to his blog or e-mail it to him.

I also prefer to do Roth conversions unless I need some money to pay expenses.
Well, at your prompting, I submitted the content as a comment a couple days ago (with a few edits), but they weren't approved. Probably too long.
livesoft
Posts: 73338
Joined: Thu Mar 01, 2007 8:00 pm

Re: Kitces on Roth Conversions vs. Harvesting Capital Gains

Post by livesoft »

petulant wrote: Fri Jul 24, 2020 4:18 pmWell, at your prompting, I submitted the content as a comment a couple days ago (with a few edits), but they weren't approved. Probably too long.
No worries, thanks. I think your pithy comments will take a while for them to digest --- maybe even 3 weeks or more. :)
Wiki This signature message sponsored by sscritic: Learn to fish.
Post Reply