Bogle along with several other well regarded investment "Gurus" have repeatedly made the point that investment returns for the typical diversified portfolio are likely to be considerably lower going forward for both stocks (given current valuations) and bonds (given current yields) and that this situation may well extend for a decade +. I am skeptical of an approach that predicates payment of large taxes now in anticipation of significant future growth of a tax free portfolio. Also, for those of us in high income tax rate states such as Oregon or California the Fed 15% rate is really more like a 25%+ when the state income tax is taken into account.
401k's, IRA's and RMD's - tax consequences
- Thread starter Ronstar
- Start date
Bogle along with several other well regarded investment "Gurus" have repeatedly made the point that investment returns for the typical diversified portfolio are likely to be considerably lower going forward for both stocks (given current valuations) and bonds (given current yields) and that this situation may well extend for a decade +. I am skeptical of an approach that predicates payment of large taxes now in anticipation of significant future growth of a tax free portfolio. Also, for those of us in high income tax rate states such as Oregon or California the Fed 15% rate is really more like a 25%+ when the state income tax is taken into account.
If it's just a one time medical event then you get to take your RMD out at 0%, and maybe some extra as well. That's a good thing. But most likely you'll be back to 25% the next year. Unless you reduced your 401k/IRA balance with Roth conversions before RMD's.
I'm not converting everything into Roth accounts. Just enough to stay pretty much in a lower tax bracket when RMD's hit. I'll still be able to take advantage of a huge deduction opportunity if it comes along.
Every Roth thread eventually gets into all these esoteric tax scenario details. While it's certainly useful and I'm learning a lot from these "robustness" discussions, in my own modeling of this dilemma, it boils down to: long-term tax savings vs foregone growth in the portfolio by paying taxes early. I don't see this discussed a lot, just the gory details on the tax side of the equation.
I'm 53, and I use a modest portfolio growth assumption of 4.4%. Yet that growth trumps everything else... even jumping high into the 25% bracket in my 70s. I've done some very detailed modeling of conversions to the top of the 15% bracket, and the result is a smaller portfolio at RIP. I've analyzed every aspect of the portfolio change, and the driving factor behind my tentative do-nothing plan is the portfolio growth enabled by continuing to defer taxes as long as possible.
Age is part of that conclusion. Plus the fact that my pension/rental income leaves precious little room for conversions. So I started modeling conversions into the 25% bracket, at various depth and timing combos. Result is usually worse; best case about the same. The one exception is some bizarre, front-end loaded withdrawal profile suggested by i-orp. Not Roth conversions, just straight withdrawals. I doubt I would ever do anything that drastic no matter what the numbers say. Layer on top of all this some healthy skepticism about tax law continuity over the next 3-4 decades, and you'll start to understand my hesitancy.
In any case, I can only assume that those early retirees without pensions (and possibly older) are able to convert larger amounts, cut deeper into RMDs, and thus offset the foregone growth with higher tax savings. Still, I have to believe there's something better than my do-nothing plan.
I'm 53, and I use a modest portfolio growth assumption of 4.4%. Yet that growth trumps everything else... even jumping high into the 25% bracket in my 70s. I've done some very detailed modeling of conversions to the top of the 15% bracket, and the result is a smaller portfolio at RIP. I've analyzed every aspect of the portfolio change, and the driving factor behind my tentative do-nothing plan is the portfolio growth enabled by continuing to defer taxes as long as possible.
Age is part of that conclusion. Plus the fact that my pension/rental income leaves precious little room for conversions. So I started modeling conversions into the 25% bracket, at various depth and timing combos. Result is usually worse; best case about the same. The one exception is some bizarre, front-end loaded withdrawal profile suggested by i-orp. Not Roth conversions, just straight withdrawals. I doubt I would ever do anything that drastic no matter what the numbers say. Layer on top of all this some healthy skepticism about tax law continuity over the next 3-4 decades, and you'll start to understand my hesitancy.
In any case, I can only assume that those early retirees without pensions (and possibly older) are able to convert larger amounts, cut deeper into RMDs, and thus offset the foregone growth with higher tax savings. Still, I have to believe there's something better than my do-nothing plan.
Just checking: Your modelling also included indexing of the tax brackets and standard deduction for inflation, right?
The growth assumption is critical to this, and so it's useful to briefly look at what happens if things turn out differently than we anticipate:
1) If growth is lower in real terms than you estimated, you might be in trouble overall. You'll need to cut back on some things. But at least you aren't getting hit with a higher tax rate, and because you didn't pre-pay a bunch of taxes up front, you have a few more dollars to see you through this rough patch.
2) If growth is higher in real terms than you'd anticipated, you are in great shape overall. Your investments are outpacing inflation by more than you thought they would when you retired, and you've got more available money to spend. In this case, is paying a higher tax especially objectionable? I don't think so.
Last edited:
Yes. Personal exemptions also.
I look at the tax rate paid for any given $ and think about it that way. In our case, we plan to delay SS to 70. We have modest pensions. Once the pension and SS are coming in, most of the 15% bracket is used up, particularly if we still have other income (dividend and LTCG). In the years prior to hitting 70, I'm planning to convert to the top of the 10% bracket (some is at 0% as mentioned thanks to standard deduction and exemption), then take LTCG/dividends to the top of the "15%" bracket, which is effectively 0%. If I use up my LTCG then I'll do more conversion. I have more money than can be converted in the time remaining in the 15% bracket, and depending on the market I may/may not have enough LTCG. If I can pay 0% and 10% rate on money at the Federal rate to convert while living off already taxed money, I'll do that all day long.
Once I hit 70, I won't have the flexibility. But if I have enough net worth and income to put me above the 15% bracket, that'll be a good problem to have.
Once I hit 70, I won't have the flexibility. But if I have enough net worth and income to put me above the 15% bracket, that'll be a good problem to have.
Cobra, your post motivated me to make some calculations based on my own finances, and I have come to a completely different conclusion. I'm not sure if it's because of our different financial situations, differing assumptions, or perhaps a mistake in my spreadsheet, since I haven't had time to analyze it carefully. Based on the Roth conversions @15% that I expect to make before RMDs begin, and assuming 4.4% growth, my spreadsheet indicates that the do nothing option will start out ahead at age 70 because of the upfront taxes I would have to pay on the Roth conversions. But the early Roth conversion option gradually catches up because of the ever increasing RMDs that get taxed at 25%. The switch-over point occurs at age 80, when the Roth conversion option pulls slightly ahead of the do nothing option. By age 90, the Roth conversion option wins by about 5x the amount of upfront taxes I had to pay in order to do the Roths. If I'm lucky enough to live to 100, the Roth conversion option would win by 15x the upfront taxes on the Roths.
Although I can't rule out a mistake in my calculations, my results make a lot of sense. The whole point of the tax laws requiring RMDs is to deplete the entire balance of one's 401k or traditional IRA over the course of one's life expectancy. You can't shelter the higher balance in the 401k forever, and once it is withdrawn you have the "pleasure" of paying Uncle Sam 25%, instead of the 15% you could have paid earlier.
My calculations also give a little more insight into how darned complicated this whole matter really is. Tax rates, life expectancy, and future investment returns all play a huge role in determining the best 401k withdrawal strategy. And none of this can be known with any certainty at all. I could get hit by a bus on my 70th birthday and lose with early Roth conversions, or live until 100 and emerge a big winner.
Karluk, thank you. I was hoping this thread would continue. In my case, the negative result from converting gets larger with age. There is no crossover point. The growth impact from paying taxes early compounds faster than the tax savings from lower RMDs and lower taxable dividends. Like you, I can't rule out an error in my spreadsheet or modeling logic. But I just spent several hours checking and rechecking everything after reading your post. The different conclusions are likely due to some combination of differences in assumptions and financial situations. However, let me ask you something... in the quote above you never actually mentioned the growth impact of the prepaid taxes. Are you only comparing RMD tax savings with the raw upfront tax dollars? Or are you looking at ending portfolio differences?
In my spreadsheet, I can breakdown the ending portfolio difference (do-nothing vs converting) at any age, into 4 components:
1. upfront taxes paid at 15%
2. growth impact from upfront taxes paid (4.4% growth assumption)
3. tax savings from lower RMDs at 25%
4. tax savings from lower taxable dividends at 15%
The kicker is #2. The taxes themselves (#1) are relatively inconsequential. Like you, my RMD tax savings exceed the raw upfront tax dollars at age 79. But the RMD tax savings never grow fast enough to exceed the growth impact (#2). So, yes, lifetime taxes are lower with conversions. But the ending portfolio is worse. And that's my acid test.
Also, in order to compare our conclusions, may I ask a few questions about the assumptions in your analysis?:
1. At what age are you starting conversions?
2. What percentage of your tax-deferred balance is ultimately converted?
3. At what age do you take SS?
4. How much of the 15% bracket do you have available for conversions?
5. What inflation rate are you assuming for tax brackets, std deduction, and exemptions?
6. Do you make Roth withdrawals? Or do RMDs still cover expenses?
Sunset
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
Ok, I'm learning some fine points here.
My above idea is a little stupid about putting the $$ into a non-ROTH.
As I missed the idea that regardless of whats invested within the Roth, its all tax-free upon withdrawl vs paying even low long term Capital Gain outside a Roth.
The metric I used for comparing strategies was total portfolio value, not taxes paid. Saving on taxes may make one feel better, but it doesn't seem to have much relevance to evaluating strategies. The real proof is in the bottom line - is my net worth higher with early Roth conversions instead of doing nothing until RMDs begin? For me, doing nothing wins until age 80, at which point Roth conversions forge a few hundred dollars ahead. At age 90, I am ahead by mid-to-high five figures, and by at 100 the difference is well into six figures. Not bad for taking some simple steps right now, although I admit it's a little disconcerting to have to wait until so late in life to see a large difference. However, for me the benefit to my heirs is also a motivation, so it's a win in my view, even if I die early.
1. age of first Roth conversion = 61. I intend to make small conversions every years from age 61-69.
2. I intend to convert a total of between 16% and 17% of the current value of my tax deferred acccount to Roth, an average of around 1.5% per year. In addition, for the purposes of the spreadsheet, I have to make additional withdrawals in order to cover the taxes due on the conversions. To convert $10,000 to Roth, for example, it's necessary to withdraw an additional $1,765 to cover the 15% Federal tax that is due on the extra $11,765 in taxable income. The total - Roth conversions plus extra money to cover taxes - is between 19% and 20% of the starting value of the deferred account.
3. SS starts at age 70.
4. The amounts converted in the spreadsheet are what take me to the top of the 15% bracket, based on current finances. Since my calculations seem to indicate that Roth conversions are a terrific idea for me, I will be looking into ways to increase the amount I can convert and still stay at 15%.
5. I don't think the inflation rate is relevant for my situation. I have a COLAed pension and SS is also COLAed. They will put me right at the top of the 15% bracket at age 70, and with the COLAs I should stay there for the rest of my life. So I assumed 25% tax on all RMDs and that I could maneuver the taxable account created from the RMDs to produce qualified dividends taxed at 15%.
6. For the purposes of the spreadsheet, I never make withdrawals from any account, except to do the Roth conversions, take RMDs, and pay taxes due. I don't expect to need it, so any spending from these accounts will be strictly discretionary.
Thanks. Sounds like we're aligned on the metric. Just a quick clarification on 4 and 5:
4. For the period from age 61-69, when you are doing conversions, can you give me some idea what percentage of the 15% bracket is available for conversions, based on other income sources? 80%? 50%? 20%?
5. For purposes of computing long-term taxes, I assume you are not leaving the current brackets, std deductions, and exemptions fixed at today's levels. They have to be indexed for inflation. I'm asking what rate you use, since that is a fairly significant assumption in my analysis. Or are you using some other approach, like stating everything in today's dollars?
Lastly, as I said before, my spreadsheet can breakdown the ending portfolio difference (do-nothing vs converting) into these 4 components.
For me at least, #1 and #4 and relatively inconsequential, and offset. Mostly, I'm curious why #2 seems to be so small in your situation vs mine, when we use the same growth assumption. The fact that you are starting conversions at 61 (vs 53 for me), may be all or part of the explanation. And if so, that could make a case for delaying conversions. But I was wondering if you could calculate that breakdown at several ages (75, 80, 85, 90), and (without providing numbers) give me some idea about the growth rate of #2 (cost) vs #3 (benefit) across that time period. Thanks again.
Would you consider posting your spreadsheet for others to see/use? (obviously without your personal data). I have used Excel quite a bit but frankly I don't think I have the talent or patience to model so many decision parameters as you seem to have done.
I am going to have to politely decline to provide that information. I'm fairly sure posting it, combined with what I posted earlier, would allow a curious onlooker to easily calculate my current income and value of my retirement account. Although I doubt that anyone really cares enough to make the calculations, I prefer not to have that information publicly available.
It simply doesn't make any difference in my calculations, because of the the COLAed nature of my retirement income and the fact that I will be near the top of the 15% bracket. If the 15% bracket increases $500 because of the effects of inflation, I expect my retirement income to also increase by $500. Similarly, a $1,000 increase in tax bracket would give me a $1,000 COLA, and so forth. I can readily imagine that this is an important factor for your spreadsheet, but not for mine.
#3 gradually overwhelms #2. The dollar amount is roughly the same as the difference in portfolio values as the Roth conversion option overtakes the do nothing option. This is what the math says should happen, and my spreadsheet confirms it. $.85 in a Roth is worth less than $1.00 in a tax deferred account only until the tax deferred money is withdrawn as part of an RMD. Then the 25% tax bite makes the Roth a winner. As an example, consider, say, $8,500 in a Roth at age 70 vs. $10,000 tax deferred at age 70. At age 90, the Roth money has compounded by 4.4% per year and is now worth $20,110 with no further taxes due. The tax deferred money has compounded to $23,659 and seems to be winning. But now an RMD mandates a withdrawal of the $23,659 and you pay 25% tax on it, leaving you with only $17,744. You have lost over $2,300 by holding the money in a tax deferred account rather than a Roth.
So to me the discrepancy in our results is likely to stem from differing RMD calculations. My numbers indicate that the RMD tables mandate ever increasing withdrawals that gradually overwhelm the 4.4% tax sheltered growth. You seem to be saying that you can shelter more of the tax deferred money longer than I think the IRS will allow.
Last edited:
Sunset
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
Funny, I had almost posted the same request, with all personal data removed of course.
growing_older
Thinks s/he gets paid by the post
- Joined
- Jun 30, 2007
- Messages
- 2,657
I do not understand what this is suggesting, and my Google search for QCD is not helping. Please explain this "current law" and QCD. Thanks.
Qualified Charitable Deduction = QCD. More here
I "get" that it's a tax efficient way to give to a charity, but I'm still trying to get my mind around how it would help a person wind up with more money to spend.
Karluk, thanks again for allowing me to pick your brain. RMD calculations are actually the most straightforward part of this entire analysis. I highly doubt that plays any role in our differing conclusions.
After several more hours of checking everything, I think I may have found the problem. My spreadsheet logic and calculations are correct, but I gave you some incorrect information. The weighted average growth assumption of my portfolio is indeed 4.4%. However, that consists of 4.1% in the tax-deferred accounts (50/50), and 5.4% in the taxable account (virtually all equities). The upfront tax is assumed to be paid from the taxable account, so the growth impact (#2 in my breakdown) is actually being calculated at 5.4%, not 4.4%. When I temporarily substitute 4.4% for the taxable accounts, the results are much more in-line with yours... breakeven at 84, then the net impact grows positive at a very fast rate.
If you might indulge me one more time just to verify this... Could you run a what-if (do-nothing vs convert) using your spreadsheet at 5.4%? If the results are similar to mine (negative result that gets bigger over time), then I think we can put this to rest. If your results are not similar at 5.4%, can you tell me what rate of return assumption is required to produce the no-go conclusion? Thanks.
This may not be entirely possible, as this is my main retirement planning spreadsheet, and is thus highly customized for our specific situation. Once I extract personal information, it may not make any sense. The Roth tab in particular is pretty disorganized right now, and not documented very well, such as column headings that only I would understand.
Let me finish validating results with karluk and then I'll give it a try.
If you not only took taxes out of a taxable account with the 5.4% growth rate, but also deposited RMDs to the same account, then the extra expected return of your portfolio gives an advantage to the do nothing option that is almost impossible to overcome, even if RMDs were taxed at a much higher rate than 25%. My estimate is the RMD tax rate would have to be around 40% for the tax advantage of Roth conversions to compensate for the lower growth rate.
If I may be permitted a short digression, I think I encountered an assumed growth assumption that caused a bias in favor of early Roth conversions when experimenting with the I-ORP calculator. I entered a higher expected return for the Roth account than the tax deferred account. That's because my Roth account is much more heavily invested in equities than my tax deferred account. I-ORP went absolutely crazy making massive Roth conversions, even putting me into the 28% tax bracket one year. This may have just been typical for how it works, but it occurred to me that it may have figured that I would continue with the higher stock composition of the Roth account, rather than rebalancing to reestablish my preferred asset allocation. As a result, the higher assumed growth inside the Roth provided a powerful incentive to get money into the account, even at a much higher tax rate.
So once I identified a likely cause for finding myself paying 28% on Roth conversions, I gave up on I-ORP. I think it's valid to assume different growth rates for different accounts, but a Roth conversion calculator needs to carefully avoid allowing asset allocation to drift after a Roth conversion in order to produce valid results.
Last edited:
W2R
Moderator Emeritus
I am among those who perceive a tremendous number of moving parts in our computations and approach to this problem. Complicating matters is the temptation to give lower lifetime tax payments a priority even higher than the amounts available to spend after paying taxes.
For those of us whose retirement is not especially early (mine was at age 61.5), most of this seems less important than I initially thought. At first I planned to not claim SS until age 70. Between SS and RMDs, I thought that would create a huge tax hit in my 70's compared with taxes in my 60's. To combat this tax situation I withdrew about twice the sustainable amount from my TSP (~=401K), using that for spending money and taking less from my taxable portfolio. The idea was to draw down my TSP considerably in order to minimize my RMD's once I hit age 70.5.
But earlier this year at age 66, only 4.5 years after retirement, I discovered that I can get divorced spousal SS while still delaying my own (still growing) SS until age 70. So, I am getting SS now, unexpectedly early. I cut back on my TSP withdrawals to a sustainable amount. I have "battened the hatches" and I am mentally ready for RMDs to hit in just 4 more short years. I projected my future TSP annual balances, determined the RMD amounts, and honestly they aren't as much as I feared they might be.
Like I said, too many moving parts! I think each of us has developed our own customized approach that we feel will best address these issues.
For those of us whose retirement is not especially early (mine was at age 61.5), most of this seems less important than I initially thought. At first I planned to not claim SS until age 70. Between SS and RMDs, I thought that would create a huge tax hit in my 70's compared with taxes in my 60's. To combat this tax situation I withdrew about twice the sustainable amount from my TSP (~=401K), using that for spending money and taking less from my taxable portfolio. The idea was to draw down my TSP considerably in order to minimize my RMD's once I hit age 70.5.
But earlier this year at age 66, only 4.5 years after retirement, I discovered that I can get divorced spousal SS while still delaying my own (still growing) SS until age 70. So, I am getting SS now, unexpectedly early. I cut back on my TSP withdrawals to a sustainable amount. I have "battened the hatches" and I am mentally ready for RMDs to hit in just 4 more short years. I projected my future TSP annual balances, determined the RMD amounts, and honestly they aren't as much as I feared they might be.
Like I said, too many moving parts! I think each of us has developed our own customized approach that we feel will best address these issues.
Last edited:
Which Roger
Thinks s/he gets paid by the post
- Joined
- Jun 5, 2013
- Messages
- 1,019
From what I can find on the web, the QCD provision expired at the end of 2013, and has not (yet) been extended by Congress, although there is speculation that it will be.
Sunset
Give me a museum and I'll fill it. (Picasso) Give me a forum ...
While we each may have developed our own approach, I am sure that some of us are simply wrong, like the music band that stayed on the Titanic while it sank.
I'm trying to learn subtle effects of actions from the discussion so I certainly value all input.
I agree the temptation to not pay much in taxes is present, sort of like the easy spending money in credit cards, but the bill of much higher taxes due to RMD + 85% of SS is the price I don't want to pay (like the final bill in credit card over-spending).
Certainly I think this affects different people to different degrees, mostly based on their tax rate when working, and their actual tax-deferred savings amount.
People who have very little in IRA/401K and no pension, don't even see this issue, they just exist on SS and maybe 5K RMD per yr.
As well someone who has amassed a huge 401K/IRA by age 60 is going to be socked hard regardless of how much is pulled out at 25% (married could only pull out ~ 1.5MM if no other income in 10 yrs, but likely has taxable dividends, so reduced a lot).
I'm trying to learn subtle effects of actions from the discussion so I certainly value all input.
I agree the temptation to not pay much in taxes is present, sort of like the easy spending money in credit cards, but the bill of much higher taxes due to RMD + 85% of SS is the price I don't want to pay (like the final bill in credit card over-spending).
Certainly I think this affects different people to different degrees, mostly based on their tax rate when working, and their actual tax-deferred savings amount.
People who have very little in IRA/401K and no pension, don't even see this issue, they just exist on SS and maybe 5K RMD per yr.
As well someone who has amassed a huge 401K/IRA by age 60 is going to be socked hard regardless of how much is pulled out at 25% (married could only pull out ~ 1.5MM if no other income in 10 yrs, but likely has taxable dividends, so reduced a lot).
Thank you for confirming this.
For my situation, holding all other assumptions constant, Roth conversions only make sense for assumed growth rates in the taxable account below 4%. Even then, the breakeven point comes so late in life (early 80s), that I still don't think I would do it. Assumed rates above 4% in the taxable account favor the do-nothing scenario. My current assumption of 5.4% is a clear no-go on conversions.
My portfolio is structured in a very conventional, tax-efficient manner. I hold mainly equities (including international), along with some real estate and muni bonds in the taxable account. This produces tax-advantaged income, with a relatively high long-term expected return. The tax-deferred accounts hold a 50/50 mix of bonds and equities, which produce high levels of interest income, with a relatively lower long-term expected return. In both cases, the conservative growth assumptions include a 1.5% haircut from the expected returns calculated by applying Portfolio Solutions' 30-year market estimates to my specific portfolio mix. I've not done anything which would intentionally bias the Roth conclusion... just using the parameters that naturally result from a rather conventional portfolio structure and AA.
As a legacy planning tool, Roth conversions still have some merit in my situation. If I measure the ending portfolio differences (do-nothing vs convert) on an after-tax basis (i.e. reduce the tax-deferred accounts by the amount of tax still owed - by someone - to the IRS), then the conversion scenario wins. But at this early point in our retirement, our overall preparedness is not so in-the-bag, that our most efficient use of cash is to start prepaying taxes for the kids. In addition, the recent proposals to require RMDs on Roths, and to require non-spousal beneficiaries to take distribution within 5 years, are not an encouraging development for Roth legacy strategies. Of course these proposals are unlikely to pass in the near-term, given the political dysfunction in Washington. But since they are clearly high on the list of revenue-generating ideas, I think it is inevitable they will pass at some point in the next 3-4 decades before any non-RMD'ed Roth account would pass to my heirs.
Unless something dramatic changes, I'm staying on the sidelines, and leaving Uncle Sam's money where I like it best... in my accounts.
truenorth418
Full time employment: Posting here.
This is the holy grail of long term retirement tax planning. An additional wrinkle for me is that my iBonds will begin to mature at age 70, adding even more taxable income to all these other considerations.
Sent from my iPhone using Early Retirement Forum
Sent from my iPhone using Early Retirement Forum
Last edited:
