The 29% Bracket in Roth Conversions

sengsational

Give me a museum and I'll fill it. (Picasso) Give me a forum ...
Site Team
Joined
Oct 13, 2010
Messages
10,766
Today I ran a bunch of "what-if" tax returns in my H&R Block 2014 tax software and determined that I hit a 29% tax rate for Roth conversions. This is with ACA PTC. Maybe 'everyone' knew that it was 29%, but I didn't. I had just been going on gut feel.

My methodology was that I kept adding $5K to my Roth conversion and then looked at the difference to my pocketbook, all inclusive with premium tax credits.

At about 223% of FPL (note that this isn't real fine-grained, so these percentages are rough), I went from zero additional cost for converting the next $5000 to a couple of hundred (tax rate of about 7%).

At about 247% of FPL, my cost rose by $1465 to convert the next $5000, which is a tax rate of 29%.

The cliff in my study came in at 415% of FPL and converting that $5K would cost me over $11K (a 223% tax rate).
 

Attachments

  • rothconvertgraph.png
    rothconvertgraph.png
    10.8 KB · Views: 62
Last edited:
So when you say a 29% "tax", you're including the increase in FIT and the decrease in ACA credits?

Just out of curiosity, what is the tax rate from 223% FPL to the top of the 15% tax bracket?
 
Last edited:
Yes, I think I paid 29% my first Roth conversion year and maybe the last couple as well. My calculations said it was worth it to avoid a 25% rate later on, though it was probably very close to a wash. The extra tax-free growth of the Roth can make up a small tax rate differential, but it's definitely diminishing returns.

Looks like I'll be converting up to my AMT limit, perhaps slightly into the ACA tax, for a few years still. Then I limit at the $250k+inflation level for a few years to stay at 25%. After that I'll be able to stay within the 15% bracket, given expected portfolio growth holds. I think I even have a few years using a state tax bracket for a limit.
 
I assume you hit the point where you filled the 15% rate and for each additional dollar added, started pushing a dollar qualified dividends or LTCGs into 15% taxable, for a 30% marginal rate. I call this the phantom 30% bracket that falls in between 15% and 25%. 30% continues until you've made all your divs/LTCGs taxable, at which point conversions fall into the 25% tax bracket.

There is no "29% bracket". In your case there is for that $5000 for one or both of two reasons:
1) you hit that point of taxing divs/LTCGs somewhere in the middle such that some of the $5000 was at 15% and most was at 30%.
2) The effect of the ACA subsidy, which is going to be unique for you. Or at least not standard for everyone.

Run your scenarios in increments of $1000 rather than $5000 in that range, than $100 and down to $1 to see this. You'll probably want to do this anyway to see exactly where the cutoff is, to avoid the 30% marginal jump.
 
I haven't looked at subsidy and tax rate verse %FPL, but from your description
At about 247% of FPL, my cost rose by $1465 to convert the next $5000, which is a tax rate of 29%.
, this is below the cliff. If one assumes family size of 2... you are well below the MAGI for the cliff and the top of the 15% bracket. So what you are likely seeing is the effect of the normal tax rate and the loss of subsidy verse income. It will change based on how much of your modeled income is LTCG and QDivs verse Ordinary income taxed items. Also tax free items will shift the subsidy as they are included in the MAGI.

If you are trying to optimize your roth conversion, be careful of looking at local effects as you may miss the global best case. In roth conversion the tax rates drop after the cliff typically up to the top of the 15% bracket. While rules of thumb are useful, these calculations really depend on individual situation.

One question, in your modeling, were you including estimated Qdivs and LTCG or just all as ordinary income?
 
I mentioned this in another thread....

The problem is that you are affecting two things at the same time....

First, you are adding in taxable income at a 15% rate...

Second, you are adding income that reduces your ACA credit at whatever rate you happen to be at... max 9.5%...


Now, you must have something else going on to get to 29%.... maybe some other credit or deduction you lose out.....
 
I think you have the 9.5% backwards. As I recall the up to 9.5% is not the subsidy, it is what you pay for health insurance after the subsidy. IOW, the subsidy is the excess of the second lowest cost bronze plan for a year over x% of your O-MAGI (which slides up as your income increases to a max of 9.5%).

So for low income the subsidy is huge as a % of income and with small income increases the subsidy lost is huge as well. Make sense?
 
I think you have the 9.5% backwards. As I recall the up to 9.5% is not the subsidy, it is what you pay for health insurance after the subsidy. IOW, the subsidy is the excess of the second lowest cost bronze plan for a year over x% of your O-MAGI (which slides up as your income increases to a max of 9.5%).

So for low income the subsidy is huge as a % of income and with small income increases the subsidy lost is huge as well. Make sense?


I guess you read my post wrong.... or it is written badly.....

Yes, this is exactly what I said.... the more income you get, the less subsidy (credit) you get... or the reverse, the more you have to pay toward your insurance... and it is a sliding scale that maxs at 9.5%....

I think it is the second lowest silver plan...
 
I think you're right... I couldn't remember whether it was silver or bronze.

I guess my point was that it would be easy to get to 29%, especially in a local where health insurance costs are high, because the subsidy as a % of income can easily exceed 14% for lower income people.
 
At about 247% of FPL, my cost rose by $1465 to convert the next $5000, which is a tax rate of 29%.

The cliff in my study came in at 415% of FPL and converting that $5K would cost me over $11K (a 223% tax rate).
Yep, it's probably a good idea to stay well clear of the cliff, at least until the very end of the year. A pop-up surprise spending requirement that forces sale of appreciated assets could get fairly expensive. Maybe that's another good reason to have some cash buffer sitting around--dry powder if the market takes a dive, and money available to buy a new roof for $10K instead of $15K+ (after tax/ACA losses).
 
Last edited:
So when you say a 29% "tax", you're including the increase in FIT and the decrease in ACA credits?
Yes. ACA credits are on the tax forms. It's money exchanged by me and the US Treasury. It's complicated to figure, but if one has anything on one of the two ACA lines of the 1040, that means you're marginal rate is not 0%, 10%, 15%, etc., but some weird other percent....like maybe 29% :)

Yes, I think I paid 29% my first Roth conversion year and maybe the last couple as well. My calculations said it was worth it to avoid a 25% rate later on, though it was probably very close to a wash.
I have no doubt you've got this nailed for your own situation, so no surprises here for the master modeler!

Just out of curiosity, what is the tax rate from 223% FPL to the top of the 15% tax bracket?
Roughly 45%. That includes the 'cliff', of course. That's going from about 20K to 70K taxable income, and going from about 50K conversion to 100K conversion. Converting that last 50K adds 22.5K to my bottom line (in lost credits and additional taxes).

I assume you hit the point where you filled the 15% rate and for each additional dollar added, started pushing a dollar qualified dividends or LTCGs into 15% taxable, for a 30% marginal rate. I call this the phantom 30% bracket that falls in between 15% and 25%. 30% continues until you've made all your divs/LTCGs taxable, at which point conversions fall into the 25% tax bracket.

That's a different animal. What I'm talking about happens at lower AGI's and has nothing to do with gapital gains.

There is no "29% bracket"....The effect of the ACA subsidy, which is going to be unique for you. Or at least not standard for everyone.
I seem to have found myself in a 29% bracket, but yes, I made no claim that others' would be exactly like me. The calculation is complicated, having to do with the cost of the second lowest silver plan, and that changes county-by-county. The point I was attempting to make was if you are getting a PPACA subsidy, you might be wise to figure out what your actual marginal "tax rate" is, not based on where your taxable income is looked-up in the tax table, but rather on the sum of the very bottom line of your 1040 plus advanced PTC.

Run your scenarios in increments of $1000 rather than $5000 in that range, than $100 and down to $1 to see this. You'll probably want to do this anyway to see exactly where the cutoff is, to avoid the 30% marginal jump.
I was manually keying this into the Block tax software, so it was labor intensive even to do at 5K increments.

I haven't looked at subsidy and tax rate verse %FPL, but from your description , this is below the cliff. If one assumes family size of 2... you are well below the MAGI for the cliff and the top of the 15% bracket.
True that. My example is a family size of 3, so even more so.

One question, in your modeling, were you including estimated Qdivs and LTCG or just all as ordinary income?
Almost everything I have is in tax advantaged accounts, so my example isn't affected by the favorable treatment of qualified dividends or long term capital gains.

Yep, it's probably a good idea to stay well clear of the cliff, ...
That's something I kinda had drilled into me from hanging out here ;) The new bit, at least for me, is that converting past a point where the subsidy gets impacted is just not a smart idea, given that I should be able to get most of my money out of tax advantaged accounts at 15%, or maybe a little 25%. So paying 29% is hard to justify only on the tax-free growth of the Roth money.
 
I am surprised it was that labor intensive....

I just copied over my 2014 tax return and deleted/added a few items....

Once I got where I thought I was going to be I just started upping one number or another... I could see the result in seconds....

I did regular income (such as conversions) and regular cap gains....


As you have found out, your marginal tax rate is much higher with regular income... and higher than most people think... that is if you have an ACA subsidy in the calculation....
 
Maybe I'm a wimp when it comes to calling something 'labor intensive', hehe! I just had the 1099-R open and kept typing in larger and larger conversion amounts. I certainly added labor by not looking only at the refund/owed amount, but I was pulling about a dozen values off of the 1040. I thought having those key values in the list might prove enlightening.
 
Maybe I'm a wimp when it comes to calling something 'labor intensive', hehe! I just had the 1099-R open and kept typing in larger and larger conversion amounts. I certainly added labor by not looking only at the refund/owed amount, but I was pulling about a dozen values off of the 1040. I thought having those key values in the list might prove enlightening.


LOL... OK... if you say that is labor intensive... :blush:


I did the same.... but wanted to look at earnings vs selling from taxable account.... just kept adding to the number and watching the top number change... wrote them down and did a quick delta....

On a few I actually did look at the return to see what was happening with them... when I was really low I was losing some credits... so I wanted to make sure I have enough earned income to get them all...
 
Back
Top Bottom