High 401K Balance, RMD's and Withdrawal Strategies

[RIGM]

My wife and I have a relatively high percentage of our assets in Tax Deferred (401k, IRA) accounts. Here is the tax status of our liquid assets, and our issue:
Tax Deferred 63%
Tax Free (Roth) 9%
Taxable (cap gains only) 28%

We face RMD's starting in about a decade that would push our marginal tax rate, and average tax rate, up substantially. So I explored several scenarios via a spreadsheet model on how best to draw down our assets in retirement: which accounts, in what sequence and by what amounts.

The model took into various assumptions on such things as inflation, asset rates of return, taxes (both current and likely reversion of post-Trump rates to prior rates), and level of spending (constrained by asset ability to support these rates and the pattern as we age), etc. As this is a impactful decision for me, I spent several days adding detail to increase the model's accuracy and, unfortunately, its complexity.

I ran 5 scenarios, from not doing any Roth conversions at all, to moderate amounts of conversion, to high amounts of conversion over the next ten years. The scenarios also looked at various sequences of withdrawal from various accounts.

For metrics for each scenario, I looked at the net present value (NPV) of the taxes paid, the annual average income, the ending estate residual value and the standard deviation of the annual average tax rates.

Without major elaboration on the model, here is what I found I should do:

- Follow the standard advice to draw down Taxable assets first and use these for living expenses. My Taxable assets should last until my late 60's.
- At the same time through my 60's, convert Tax Deferred assets to Tax Free. Do this to (1) reduce the Tax Deferred balance and (2) to smooth tax rates over the span of retirement. Without Roth conversions at a moderate level, my annual tax rates in my 60's are low (capital gains only, minimal income taxes) and my RMD's later are high; I should take advantage of the low income tax rates in my 60's to reduce my taxes in my 70's and beyond (this reduces the lifetime NPV of my taxes).
- Post age 70, draw RMD's from my Tax Deferred accounts for living expenses. Assuming a long life and drawing each year the RMD, the Tax Deferred assets will last about as long as I will live.
- Post age 70, supplement my income as needed from my Tax Free assets. This keeps my annual tax rate lower than if I took even more than my RMD's from my Tax Deferred account. If I only need the RMD's to live on, the Tax Free account can remain untouched without any tax consequences, but if I need more, the source of additional funds should be my Tax Free account so I don't raise my tax rate significantly later in life.

I admit I haven't explored all the posts on this site on RMD strategies, and so this conclusion might be nothing new to some here. The advice I've seen in other posts is generally to keep the annual average tax rate about level. That is solid advice. That occurs for me in the plan above, but I also found I could achieve a level tax % with other, less optimal drawdown patterns (e.g., spending both Taxable and Tax Deferred each year throughout my entire life). So I needed to explore the specific accounts that I would tap, and the order, to maximize my benefit beyond just leveling my tax rate across time.

There are the usual caveats:
- Nobody can predict the future and so this is at best a SWAG, sophisticated wild ax guess. Changes in assumptions or conditions may affect the conclusion.
- Each situation can be unique, and so my results may not apply to all. For example, I have no immediate heirs and so am not concerned about positioning my accounts for heir tax consequences (although most residual value in this plan is in a Roth). Also, the amount of draws from each type of account likely needs to be tailored individually to the amount of money available in each type of account, and the tax pain being faced by each person.

Still, for those with large Tax Deferred accounts, the above plan might be a good starting place to explore your own situation.

[Spock]

You've obviously put a lot of thought into this, so I suspect you've already factored it in... I only bring it up since I don't see it mentioned: Be sure to figure in the increased taxability of your SS benefits aka "the tax torpedo". Above some very low thresholds 85% of your SS benefit is considered taxable income. It makes the Roth conversions a no-brainer and in some cases justifies a step up in tax brackets (ex between 10% and 12% brackets).

[RIGM]

I am toast on the SS tax issue no matter what I do short of dying early. What I didn't mention above, however, is that in all scenarios I am deferring taking SS to age 70 so that I have more "room" to do Roth conversions and not have the tax effect from the SS income during my 60's.

[Clone]

One thing that I did with mine was to look at some 'what if?' scenarios. What if one of us dies at age 70? What does that do to the tax situation. Also, if both of us pass, what appears to be the tax situation for our heirs (not a problem for you, but could be for others). I looked at what incremental tax rate is in each of those situations, and conclusions were to convert aggressively early.



What was interesting in I-ORP was the impact of inflation. Your SS and any COLA income will ratchet up the taxable income. Thus reinforcing the early conversion plan. Folks that are managing income for ACA subsidies are very limited in how much they can convert.

[street]

Thanks for sharing your data on RMD. I will be in a world of hurt when RMD comes around in a decade also. I am not taking out any right form 401K or IRA to keep my HL premium as low as I can. SO saving there but will pay for it later. I plan as soon as I hit 65 I will start taking money from tax deferred accounts in hope to draw the down in hopes I will be not have to be in the 32% tax bracket. I don't beleive that is going to happen. The reason is those accounts are going to grow, I'm assuming over the next 10 year or so.

So, did you looked at taking a large amount out from your tax deferred accounts, for one year and pay the high tax bracket?? Pay the taxes and get it over with and start growing that money without taking it every year as a RMD? If so how bad of a plan would that be??
Thanks

[ziggy29]

In a situation like this I might be inclined to withdraw as much as I can in lower tax brackets in order to reduce future RMDs which are likely to be taxed at higher levels.

[RIGM]

Street, yes, I looked at that, and the best plan seems to level the tax burden across all years of retirement. This keeps more of your money in lower brackets. If you do a big conversion or withdrawal from a Tax Deferred account, even in just one year, then a larger chunk of that money will be taxed at a higher rate. This raises the total tax paid and the lifetime average tax rate as compared to moderating the conversion over a period of years.

Now, what the marginal and average tax rate will be when leveling taxes across all years will depend mostly on the amount of wealth and the amount withdrawn / converted each year. So if someone is very wealthy, they might be getting taxed with an average rate above the 24% or 32%, and trying to keep more money out of the higher brackets. Someone with more modest wealth / spending might have an average above the 12% or 22% brackets and trying to keep more money out of the next higher bracket.

I started thinking of the curve of the yearly average tax rate as a balloon. If I squeeze it to be low in my 60's (keeping my taxable income low), then that puts upward pressure on the rates I pay later in life. Similarly, you could take a large conversion early, and that would put downward pressure on your rates later in life, but again, the best seems to be a smooth rate across retirement.

That is, unless you have a concern that promotes you doing an early and large conversion: for example, worried about being affected by high taxes if they go back up to 70% or the establishment of a wealth tax, etc. The odds of those occurring, while not zero, seem to me to be remote (more probable for the billionaires, however).

For my plan, I assumed I would spend an amount that would draw down my liquid assets by about 80% over a long life. In reality, I may never spend that amount, as it is higher than I'm used to spending, but that was the assumption. Then, I looked at where that money would come from - which accounts and when. That was the piece of the plan that had some effect on the total taxes I paid over my life.

[RIGM]

Quote:

Originally Posted by ziggy29 In a situation like this I might be inclined to withdraw as much as I can in lower tax brackets in order to reduce future RMDs which are likely to be taxed at higher levels.

I agree, and the key issue is how much now versus later, and from which accounts, to minimize the lifetime total tax burden. See my "balloon" comment in my previous reply.

[RetireAge50]

I am in a similar tax situation with 69% tax deferred. I would like my overall spending throughout my life to be approximately equal and then die broke. My tax approach originally was to use my tax deferred accounts for about 69% of my portfolio withdrawals each year. This would result in me using most or all of the lower brackets every year (and then the remainder would be in the upper brackets).

I have had to modify my thinking a little bit as the ACA subsidies get lost if I use this much tax deferred money. So my new strategy for now is to only use tax deferred money in the low brackets and use other money for the remainder of spending.

[googily]

It sounds like you have looked at this from all the angles, except, as one person said above, what happens if one of you dies, when the taxes for the remaining spouse will be a bigger bite, especially at RMD time (which is in essence a vote for aggressive Roth conversions). If you've saved well enough for two to retire comfortably, that's a big chunk of money for one person alone, with much worse tax rates.

I'm still working at 52 (widow, no kids) my base fed bracket is 22%, and I am converting well into the 24% bracket, along with maxing out my mega backdoor Roth. I only opened a Roth last year, so I have a lot of catching up to do.

My own crazy spreadsheet and lots of scenarios show that it is best for me to convert like this now, and even into the 25% bracket after rates presumably revert, because the combination of that and withdrawing from my tax-deferred through my 60s will bring my RMDs down to a much lower level. My rate would probably be 28% in my 60s but then down to 25% in my 70s. If I don't convert, I'd get up into 33% from 70 onward.

Watch out with the spreadsheet--it can become an addiction. As I now know!

[GalaxyBoy]

I'd be interested as to how your analysis compares to the results of using I-ORP.

[Dtail]

85% TIRA for us. As others have mentioned, I am maximizing ACA tax subsidies instead of Roth conversions until 65 y.o.
For us, there is also the aspect of not spending down the TIRA too much before SS at 70 y.o.
A whole balancing act.

[RIGM]

Clone, The other thing to factor in is that income tax rates are indexed to inflation. So while SS, COLA-impacted income and assets may increase in value, it is the rate above inflation that matters.

Related to that, and as an aside, the online RMD calculators show having to take big distributions late in life, but the assumed rate of growth in the calculation needs to be after inflation since the tax brackets will change at that rate over time. I initially thought I was going to be in the 37% marginal bracket in my 80’s until I realized this. It is a whole different situation at 2-3% real return vs 5-6% over 10 years.

I’m in excellent health, and while any one of us can die at any time, I have good odds at a long life. If I do die early, the RMDs for my naturally frugal wife won’t be an issue. She will have plenty of assets to live beyond age 100. I’m looking at the most probable scenario: being around well past 80.

I’m lucky that we have retiree healthcare through my wife’s past employer. I can’t comment on the effect of the ACA on this planning.

Like others here, I too am naturally inclined to increase conversions now because, hey, who knows what is coming. I initially was going to convert much more each year, but I now will moderate that somewhat.

I haven’t used I-ORP, but guess I should take a whack at it.

[street]

RIGM thanks for the response. I just dread the thought of what to do but most likely won't do much till I'm 65 to start drawing out of deferred accounts. It will be ugly and right now it looks like I would be in the 32% bracket and many years to go for that money to grow. I guess when the time gets closer I will need to come up with the best plan that I can.

I never have understood why a person with more money has to pay more taxes. It almost seem discrimination and unfair. I'm not against having to paying taxes and still a great deal to be able to delay taxes but being penalized for having more money isn't a fair thing in my books.

[RIGM]

So I just used the basic I-ORP (I haven't explored the advanced version yet). I need to look at this a bit more, but my initial conclusion is that I'm not a fan. Here's why:

- Biggest reason is that I-ORP completely excludes any Roth conversions. This is the big question I was trying to answer. As a justification, they state that a referenced study says that conversions "offer little economic advantage," but the study itself states conversions reduce tax by 19% and increase disposable income by 1%. How is that not advantageous? Maybe for the average person with the average Tax Deferred account, the conversion doesn't matter. Then again maybe it does, even for them, with lower tax rates in effect now. Certainly with 63% of my assets in Tax Deferred status, it has to be a matter considered in my planning.
- It plans for a life expectancy of 92. Joint life tables show a considerable probability that one or the other spouse will live longer than that. I just looked at a joint life probability table: for a "white collar" man and woman, both 60, there is a 53% probability that at least one will be alive at 93. It seems age upon death should be a configurable value in I-ORP since we all have different genes.
- It assumes you sell your house at age 80. Huh? The age thing and the configurable thing again.
- It assumes 2% inflation. That is what I used (and even then we both could be wrong - no way to do any sensitivity analysis in I-ORP on that).
- It assumes a 4% increase in spending in retirement. I have not read the referenced study, but there is other research out there that indicates spending declines significantly over time in retirement. I assumed in my model that spending would be constant, adjusted for inflation.
- It assumes a 7% equity rate of return and 3.5% rate of return on bonds. I believe that in this market, that is too high. I used a blended rate of return of 2% above inflation, both because the market is high (CAPE index) and to be conservative.

I will look at the advanced version of I-ORP, but I'm initially not impressed.

[CardsFan]

Quote:

Originally Posted by RIGM So I just used the basic I-ORP (I haven't explored the advanced version yet). I need to look at this a bit more, but my initial conclusion is that I'm not a fan. Here's why: - Biggest reason is that I-ORP completely excludes any Roth conversions. This is the big question I was trying to answer. As a justification, they state that a referenced study says that conversions "offer little economic advantage," but the study itself states conversions reduce tax by 19% and increase disposable income by 1%. How is that not advantageous? Maybe for the average person with the average Tax Deferred account, the conversion doesn't matter. Then again maybe it does, even for them, with lower tax rates in effect now. Certainly with 63% of my assets in Tax Deferred status, it has to be a matter considered in my planning. - It plans for a life expectancy of 92. Joint life tables show a considerable probability that one or the other spouse will live longer than that. I just looked at a joint life probability table: for a "white collar" man and woman, both 60, there is a 53% probability that at least one will be alive at 93. It seems age upon death should be a configurable value in I-ORP since we all have different genes. - It assumes you sell your house at age 80. Huh? The age thing and the configurable thing again. - It assumes 2% inflation. That is what I used (and even then we both could be wrong - no way to do any sensitivity analysis in I-ORP on that). - It assumes a 4% increase in spending in retirement. I have not read the referenced study, but there is other research out there that indicates spending declines significantly over time in retirement. I assumed in my model that spending would be constant, adjusted for inflation. - It assumes a 7% equity rate of return and 3.5% rate of return on bonds. I believe that in this market, that is too high. I used a blended rate of return of 2% above inflation, both because the market is high (CAPE index) and to be conservative. I will look at the advanced version of I-ORP, but I'm initially not impressed.

You need to use the advanced version to use Roth conversions. Also, you can vary many of the inputs that are pre-selected in the basic version.

[RIGM]

Ok, thanks. I will look at that.

[Floridaman10]

Quote:

Originally Posted by street RIGM thanks for the response. I just dread the thought of what to do but most likely won't do much till I'm 65 to start drawing out of deferred accounts. It will be ugly and right now it looks like I would be in the 32% bracket and many years to go for that money to grow. I guess when the time gets closer I will need to come up with the best plan that I can. I never have understood why a person with more money has to pay more taxes. It almost seem discrimination and unfair. I'm not against having to paying taxes and still a great deal to be able to delay taxes but being penalized for having more money isn't a fair thing in my books.

Street, very good point that most miss. I always have thought charging higher percentages for making more is unfair.

[target2019]

Quote:

Originally Posted by RIGM So I just used the basic I-ORP (I haven't explored the advanced version yet). I need to look at this a bit more, but my initial conclusion is that I'm not a fan. Here's why: - Biggest reason is that I-ORP completely excludes any Roth conversions. This is the big question I was trying to answer. As a justification, they state that a referenced study says that conversions "offer little economic advantage," but the study itself states conversions reduce tax by 19% and increase disposable income by 1%. How is that not advantageous? Maybe for the average person with the average Tax Deferred account, the conversion doesn't matter. Then again maybe it does, even for them, with lower tax rates in effect now. Certainly with 63% of my assets in Tax Deferred status, it has to be a matter considered in my planning. - It plans for a life expectancy of 92. Joint life tables show a considerable probability that one or the other spouse will live longer than that. I just looked at a joint life probability table: for a "white collar" man and woman, both 60, there is a 53% probability that at least one will be alive at 93. It seems age upon death should be a configurable value in I-ORP since we all have different genes. - It assumes you sell your house at age 80. Huh? The age thing and the configurable thing again. - It assumes 2% inflation. That is what I used (and even then we both could be wrong - no way to do any sensitivity analysis in I-ORP on that). - It assumes a 4% increase in spending in retirement. I have not read the referenced study, but there is other research out there that indicates spending declines significantly over time in retirement. I assumed in my model that spending would be constant, adjusted for inflation. - It assumes a 7% equity rate of return and 3.5% rate of return on bonds. I believe that in this market, that is too high. I used a blended rate of return of 2% above inflation, both because the market is high (CAPE index) and to be conservative. I will look at the advanced version of I-ORP, but I'm initially not impressed.

As mentioned, the expert version gets you additional options.
The home sale option--I skipped that as part of my quest to bring i-ORP down to earth.

[Marc]

One other thing to look at is IRMAA. Even with indexing starting next year, I think we will be just over the IRMAA cliff when we start Medicare in 2022/2023. Although I aim to stay right below the top of the 22% bracket (income for living and any ROTH conversions), I am thinking I might have to move up to 24% bracket in alternating years just to halve the impact of IRMAA; I think it will be around $270 per month for the two of us if we go over the cliff.

Marc