Delaying SS OR take IRA distributions early

[clifp]

In the recent thread on retiree paying zero in taxes. Onward posted a very interesting link to a Rick Ferri piece Avoiding Taxes As A Retiree Isn’t So Easy. Frankly it was an eye opener for me.

The article is well worth reading, but the TLR is this.
Even if you have been happily enjoying modest tax rates in the early stages of retirement, when you turn 70.5 RMD are going substainially increase your taxes. The example Rick uses is an affluent couple who retires with $2 million in 401K/IRA at age 60,by age 70.5 2 million has grown to $4 million with a 7% return. They are required to withdraw at least 144k, putting them in pretty high tax bracket and decreasing a number of other benefits.It is worth noting that if you have a $1 million in your retirement accounts by 50, you will also likely hit $4 million by age 70.5.

What Rick fails to mention is if you followed conventional wisdom you also delayed taking social security (for at least one spouse) until age 70. Because SS benefit increase pretty rapidly a working couple that take SS at age 70 can easily get another $50-70K in SS benefits. This means that you'll see a spike in income of more than $200K for a couple when they turn 70 and the tax bracket will shoot up from say 15% to 33%.

So here is the interesting dilemma. Conventional wisdom is you withdraw money from tax deferred account last to maximize the benefits of tax free compounding. Conventional wisdom is that you delay SS, because this is the least expensive COLA adjusted annuity you can purchase.

I am pretty sure that conventional wisdom is correct in both cases if these were strictly independent decisions, but they are not. If you do both you'll end up with a big spike in income at age 70. I haven't done any calculations, but my estimate is that tax wise you are much better off smoothing your income in your 60s than seeing a big spike in the 70s.

The question is which is the better approach, withdrawing money from your IRA and/or doing a ROTH conversion or taking SS at full retirement age or possible even at 62. Or even more fundamentally how do you go about modeling this question?

 

[Walt34]

I am pretty sure that conventional wisdom is correct in both cases if these were strictly independent decisions, but they are not. If you do both you'll end up with a big spike in income at age 70. I haven't done any calculations, but my estimate is that tax wise you are much better off smoothing your income in your 60s than seeing a big spike in the 70s.

That's the direction we're leaning toward, taking out from a 457 (basically the same thing as a 401k) and put off SS for the later higher benefit. Admittedly we haven't fully done the math yet. But with me stopping work next month, that means next year we'll never again be in such a low tax bracket since it'll go up with SS income and it makes sense to withdraw from the 457 then and keep us under the 25% bracket.

 

[Jack_Pine]

We are leaning towards converting standard IRS to Roths some at a time. How much, have not decided.

I-ORP has this happening while you are drawing down post tax account. I am not sure how they come up with the amount to convert each year, though.

 

[RockyMtn]

Once RMD's, SS and Pension payments begin it is hard to turn off the spigot. Your taxes are going to be what they are going to be!

My plan is to convert IRA's, 401k's to Roth's during those years prior to SS and RMD's taking effect. This way I can control withdrawals to insure the most tax efficient drawdown strategy versus the IRS doing it for me. If you combine that strategy with the strategy of muni's, etf's, etc in after tax accounts mentioned in the "How Retirees Pay Zero Taxes" you can create an income smoothing effect that will help minimize (although not eliminate) taxes as you age.

 

[Chuckanut]

The example Rick uses is an affluent couple who retires with $2 million in 401K/IRA at age 60,by age 70.5 2 million has grown to $4 million with a 7% return.

If I could be guaranteed my current assets would double in 10 years, I would be very happy to pay the extra taxes.

Alas, no such guarantee exists at today's rates of return.

Like many I find the taxes/SS/Roth-Conversion planning to be quite complicated. It really depends on one's assumptions of things like future tax rates, and how the government will 'price' certain services. My current calculations show that for MY scenario, converting regular IRA's to Roths is pretty much a wash in terms of the amount of money I will have left after taxes. But, I can't help but think that the trend towards government using income to figure our costs for certain services will continue into the future. So, the Roth conversion seems sensible even though it will somewhat deplete my asset base today.

As much as I hate to pay taxes earlier than necessary, some Roth conversions seem to make sense to me, at least to the point where my Roth/RIRA ration changes from about 1/4 to maybe 1/2. Must my 2 cents.

For those who wish to calculate their expected MRD:

https://web.fidelity.com/mrd/application/MRDCalculator

 

[packrat44]

We have a situation that will cause a spike in our income/taxes sometime after I turn 70. Therefore, our plan is to convert tIRAs to ROTHS from age 62 to 70. Also, delay my SS until 70 to maximize our conversions at a lower tax rate (and later, for the spousal survivor benefits). DW with draw SS at 62 (she is about 9 years younger). Having a couple of modest pensions and not having high medical insurance costs (retired military) keeps from having to dip heavily into the retirement funds, with the exception of using some funds from our post-tax account to cover the conversion costs.

The Tax Man is like the Fram Filter man - "You can pay me now, or you can pay me later". For some it is more beneficial to pay early, for others it is more beneficial to pay later. But in the end, we all have to pay the Tax Man.

 

[Bikerdude]

But in the end, we all have to pay the Tax Man.

"Now my advice for those who die
Declare the pennies on your eyes
Cos I'm the taxman, yeah, I'm the taxman
And you're working for no one but me
Taxman!"

 

[Animorph]

Our current, and I believe finally final, plan is to Roth convert just up to the point where we would have to start paying AMT, through 2019. That's well into the 25% tax bracket for most years and 28% this year. AMT is bad when you're in the exemption rollback region, where your marginal rate is much higher than the nominal rate. 2020 is kind of a transition year, halfway between 2019 and 2021. For 2021 and beyond we will fill up the 15% tax bracket.

I'm using the current tax rates with inflated brackets. DW and I will be taking SS at 70, with some spousal benefits for a few years. All very specific to our numbers. I believe that is our optimum scenario for maximizing yearly after-tax spending, with a specified end portfolio value and simplistic investing and inflation numbers.

I think the main takeaway is that you probably want to maximize Roth conversions as early as possible, to a specific AGI, tax bracket, or AMT level that approximates your potential RMD tax bracket and preserves nearby tax credits or subsidies. The Roth has the advantage of holding more after-tax value than an IRA for the same amount of account dollars, so the sooner you can do the conversions the better.

 

[packrat44]

Our current, and I believe finally final, plan is to Roth convert just up to the point where we would have to start paying AMT, through 2019. That's well into the 25% tax bracket for most years and 28% this year. AMT is bad when you're in the exemption rollback region, where your marginal rate is much higher than the nominal rate. 2020 is kind of a transition year, halfway between 2019 and 2021. For 2021 and beyond we will fill up the 15% tax bracket.

I'm using the current tax rates with inflated brackets. DW and I will be taking SS at 70, with some spousal benefits for a few years. All very specific to our numbers. I believe that is our optimum scenario for maximizing yearly after-tax spending, with a specified end portfolio value and simplistic investing and inflation numbers.

I think the main takeaway is that you probably want to maximize Roth conversions as early as possible, to a specific AGI, tax bracket, or AMT level that approximates your potential RMD tax bracket and preserves nearby tax credits or subsidies. The Roth has the advantage of holding more after-tax value than an IRA for the same amount of account dollars, so the sooner you can do the conversions the better.

We have been converting to the top of the 25% bracket each year. However, we had too much LTCG 2011 tax year and got hit with a sizable AMT expense. I was watching the front door and the Tax Man slipped in the back door. AMT hurts.

This year we have cut back to maxing the 15% bracket. The following year we will complete our conversions and I will turn 70 and start receiving SS. Six months later DW will start her SS benefits at 62.

 

[omni550]

We have been converting to the top of the 25% bracket each year. However, we had too much LTCG 2011 tax year and got hit with a sizable AMT expense. I was watching the front door and the Tax Man slipped in the back door. AMT hurts.

As someone who has a large tIRA and is looking down the road at significant RMDs at 70.5+, I am most interested in learning more about tax reduction strategies.

Do you know if Turbotax will 'predict' whether you will trigger the AMT when trying to determine what a "reasonable" amount to convert from tIRA to Roth?

omni

 

[youbet]

Here's my post from the "How Retirees Pay Zero Taxes" thead. It seems more appropriate here than there.

DW and I maxed 401k, 403b and TIRA's for many years. We were not eligible for direct Roth contributions and this was in the pre-backdoor time period. Some of the TIRA's were after tax contributions. Had I known how the investment returns and tax laws were going to work out, I would have toned the deferred contributions down a bit and put that money into tax managed investments.

Since retiring (11 yrs DW and 7 yrs me), we've deferred one of our pensions and used that space to do some Roth conversions. We didn't get on this right away (only 4 yrs ago) and I wish we would have started sooner. Once the second pension and RMD's kick in, we'll be in a higher tax bracket than the single pension plus Roth conversion years.

We'll have very little tax management flexibility once RMD's start in 5 - 6 years. Pension + RMD will be more than our budget so we'll actually be paying tax on ordinary income of more than we spend.......

We're doing fine and have been enjoying a comfortable middle class lifestyle in our FIRE years. No complaints. We'll pay whatever tax we have to pay and go on. I only mentioned our scenario since it is similar (just smaller numbers) to the scenario in the article posted by Onward in post #125. There will be no tax holiday for us in retirement........ And I think a lot of other folks on this board as well.

 

[RunningBum]

http://research.prudential.com/docum...ityNov2012.pdf was posted in another thread and makes a case for how to juggle your withdrawals to prevent that higher SS taken at age 70 from being a higher tax burden. I haven't studied it in detail enough to fully understand it but I believe the concept is to take some other income between 62-70, then the higher SS doesn't bite as much.

 

[walkinwood]

Rick Ferri's piece doesn't tell the whole story.

I think he should be comparing two (or more) scenarios - one where a person starts withdrawing from IRAs at 59 1/2 (60, 65 etc) and the other where they wait till RMDs are mandated. Which scenario leaves you with more total cash after taxes?

I don't know what he'll find, but if someone knows of such a study, it will be more useful than this article. Add taking SS (at various ages) to the mix and you may have something really useful.

Btw, it is also being discussed over at bogleheads
http://www.bogleheads.org/forum/viewtopic.php?f=10&t=117820&sid=4eb04f3d0e0d50a8e92dc6fe00f1ec22

 

[Which Roger]

I've been thinking about this, too. The Piper book (from last week's free E-books thread) provides a lot of insight into the trade-offs when it comes to retirement tax planning.

Managing taxes is important, but we need to also keep in mind that it's only a means to an end, that end being making sure we don't run out of money before we die. If I find myself in the top tax bracket when I turn 70 1/2, it means I've won the game in a landslide, and can pat myself on the back while I write a big check to the IRS.

There is a significant correlation between minimizing taxes and portfolio survival, but they don't go in lock step. Also, I can't assume that paying less in taxes now will mean I pay more in the future. I can think of at least 3 reasons for this:
- My portfolio might perform very poorly and my assets will be nearly depleted.
- Contrary to a lot of what is often assumed, tax rates may be significantly lower in the future
- I will be dead and therefore in the zero percent bracket

Given this uncertainty, a dollar that I know I will pay in taxes this year needs to be weighted more than a dollar that I might pay in the future.

I realize that I've omitted some variables (especially spouses and heirs), but I hope I've expressed the general idea clearly.

 

[packrat44]

As someone who has a large tIRA and is looking down the road at significant RMDs at 70.5+, I am most interested in learning more about tax reduction strategies.

Do you know if Turbotax will 'predict' whether you will trigger the AMT when trying to determine what a "reasonable" amount to convert from tIRA to Roth?

omni

This would not have happened if I had used "due diligence" and entered my projected LTCG (prior to actually selling), along with projected conversion amount and projected other income into a practice TurboTax return. I fell asleep at the wheel.

Figure out what you would have to convert yearly to get to your desired target when RMDs would start. Enter that yearly conversion amount along with your other pertinent data into a practice TurboTax return. See how the numbers shake out and adjust your conversion accordingly.

 

[Independent]

I did some calculations when I was 62. Pro forma tax returns with SS and RMDs and pension, for multiple future years. For us, the only clear winner was to do some rIRA=>Roth transfers when our tax rates were low. Other decisions seemed to be mostly a wash.

Note that the factors in the SS taxability test aren't indexed for inflation. That had an impact on us, because it looked like 85% of our SS income would be taxable in almost all situations.

 

[packrat44]

- Contrary to a lot of what is often assumed, tax rates may be significantly lower in the future
.

Sorry about that, the made me do it. It is possible, but not probable considering how our country is increasing our debt load.

 

[RunningBum]

There is a significant correlation between minimizing taxes and portfolio survival, but they don't go in lock step. Also, I can't assume that paying less in taxes now will mean I pay more in the future. I can think of at least 3 reasons for this:
- My portfolio might perform very poorly and my assets will be nearly depleted.
- Contrary to a lot of what is often assumed, tax rates may be significantly lower in the future
- I will be dead and therefore in the zero percent bracket

Given this uncertainty, a dollar that I know I will pay in taxes this year needs to be weighted more than a dollar that I might pay in the future.

On the other hand:
- My portfolio might perform just fine and generate more income in the future, pushing me into a higher bracket and/or more taxable SS benefits
- Tax rates may follow the more common assumption that they will increase
- Means testing for various benefits in the future may favor those with lower balances, and having taxes paid gives me a lower balance without necessarily having a poorer position.
- I may still be alive, in which case I'm dealing with the ramifications of past decisions, rather than being dead and having no concern with what I've decided.

Still a lot of uncertainty, but if I was to base weights on the likelihood of which events will happen, the ones in my post seem more likely, at least to me. YMMV.

 

[clifp]

I've been thinking about this, too. The Piper book (from last week's free E-books thread) provides a lot of insight into the trade-offs when it comes to retirement tax planning.

Managing taxes is important, but we need to also keep in mind that it's only a means to an end, that end being making sure we don't run out of money before we die. If I find myself in the top tax bracket when I turn 70 1/2, it means I've won the game in a landslide, and can pat myself on the back while I write a big check to the IRS.

There is a significant correlation between minimizing taxes and portfolio survival, but they don't go in lock step. Also, I can't assume that paying less in taxes now will mean I pay more in the future. I can think of at least 3 reasons for this:
- My portfolio might perform very poorly and my assets will be nearly depleted.
- Contrary to a lot of what is often assumed, tax rates may be significantly lower in the future
- I will be dead and therefore in the zero percent bracket

Given this uncertainty, a dollar that I know I will pay in taxes this year needs to be weighted more than a dollar that I might pay in the future.

I realize that I've omitted some variables (especially spouses and heirs), but I hope I've expressed the general idea clearly.

You make some very good points. The eye opener for me is I only need to get to get 6.1% return in my IRA to hit 4 million, in the 10 years from Dec 2002 to Dec 2012 I got 5.9%, its up 17% this year. Even assuming 3% inflation when I add in my SS at 70, I'm looking at 150% increase in my income at 70 from today. As you say I'll have won the game.

But besides the huge increase in taxes, I think I'd be better off spending more money in my late 50s and 60s. One obvious way is to do Roth conversion more aggressively. But like you I have weighted the tax dollars I have to pay today much more than tax dollars I might have to pay in the future.

One of the things we do know is that tax brackets are indexed for inflation.
So Rick and I both probably both exaggerated the impact by using nominal gains, and not adjusted the tax brackets for inflation.

A big factor of course is future returns. One of the things that I have been trying to estimate is the historical likelyhood of various portfolio returns.

E.G. A 50/50 portfolio will have a real return over 10 years
90% of the time at least 1%
75% at least 2%
50% at least 3%
25% at least 5%
10% at least 7%

I tried playing with FIRECalc but I couldn't figure out how to get the answers I was looking for.

 

[haha]

In the recent thread on retiree paying zero in taxes. Onward posted a very interesting link to a Rick Ferri piece Avoiding Taxes As A Retiree Isn’t So Easy. Frankly it was an eye opener for me.

The article is well worth reading, but the TLR is this.
Even if you have been happily enjoying modest tax rates in the early stages of retirement, when you turn 70.5 RMD are going substainially increase your taxes. The example Rick uses is an affluent couple who retires with $2 million in 401K/IRA at age 60,by age 70.5 2 million has grown to $4 million with a 7% return. They are required to withdraw at least 144k, putting them in pretty high tax bracket and decreasing a number of other benefits.It is worth noting that if you have a $1 million in your retirement accounts by 50, you will also likely hit $4 million by age 70.5.

What Rick fails to mention is if you followed conventional wisdom you also delayed taking social security (for at least one spouse) until age 70. Because SS benefit increase pretty rapidly a working couple that take SS at age 70 can easily get another $50-70K in SS benefits. This means that you'll see a spike in income of more than $200K for a couple when they turn 70 and the tax bracket will shoot up from say 15% to 33%.

So here is the interesting dilemma. Conventional wisdom is you withdraw money from tax deferred account last to maximize the benefits of tax free compounding. Conventional wisdom is that you delay SS, because this is the least expensive COLA adjusted annuity you can purchase.

I am pretty sure that conventional wisdom is correct in both cases if these were strictly independent decisions, but they are not. If you do both you'll end up with a big spike in income at age 70. I haven't done any calculations, but my estimate is that tax wise you are much better off smoothing your income in your 60s than seeing a big spike in the 70s.

The question is which is the better approach, withdrawing money from your IRA and/or doing a ROTH conversion or taking SS at full retirement age or possible even at 62. Or even more fundamentally how do you go about modeling this question?

I really don't know. One thing jumps out from your post- it seems to me that if taking money from 401K or Trad IRA is in the plan, generally it will be much better to do Roth conversions than just withdrawals. Same tax in either case, but the Roth conversion sets you up better for the future. I never had this early retirement tax holiday that people talk about, partly because while I have been separated for many years, I didn't get divorced until later, so MFJ was our status and by then wife was making pretty good money.

I did some Roth conversions after divorce but before I started RMDs or ss, but the marginal rate on these was always 25%. I stopped at the top of this, or the 85K which is the limit for paying base Medicare rate. Recently the $85,000 is the stop point. I think it was likely worthwhile anyway, as I have a better % return in this Roth account since I do not fear huge taxes on a big win, and of course if I ever take any money out it will be tax free.

I do think that it will be unlikely that a million dollar TIRA in 2010 will become a 2 million dollar IRA 10 years later, but anything is always possible. Today I tend to prefer using my 15% bracket on 0% taxation of Qualified dividends and LTCG, to the extent that I have some room. For a single taxpayer getting SS and RMDs, that 15% bracket is close to full before you even start tinkering, and for some it will be beyond full.

Ha

 

[haha]

Contrary to a lot of what is often assumed, tax rates may be significantly lower in the future

If you have time, could you make a case for this viewpoint?

I think it is impossible, but I'm all eyes for your ideas.

Ha

 

[M Paquette]

So here is the interesting dilemma. Conventional wisdom is you withdraw money from tax deferred account last to maximize the benefits of tax free compounding. Conventional wisdom is that you delay SS, because this is the least expensive COLA adjusted annuity you can purchase.

I am pretty sure that conventional wisdom is correct in both cases if these were strictly independent decisions, but they are not. If you do both you'll end up with a big spike in income at age 70. I haven't done any calculations, but my estimate is that tax wise you are much better off smoothing your income in your 60s than seeing a big spike in the 70s.

The question is which is the better approach, withdrawing money from your IRA and/or doing a ROTH conversion or taking SS at full retirement age or possible even at 62. Or even more fundamentally how do you go about modeling this question?

I didn't try modeling it, beyond looking at what would happen if I left my IRA untouched and delayed SS to age 70 as longevity insurance (for my wife; remember how benefits interact.) With it untouched, and adding in distributions from taxable accounts, SS was heavily taxed. If I tapped the IRA from 59 1/2, then with taxable distributions I could still keep taxable income quite low, run down the IRA, and keep most of SS tax-free past age 70.

So that's the plan for me.

 

[Which Roger]

If you have time, could you make a case for this viewpoint?

I think it is impossible, but I'm all eyes for your ideas.

Ha

I don't think I can make a convincing case either way, but my point is that the common wisdom (which is currently that tax rates will be higher) isn't always right. As Yogi Berra said, "it's hard to make predictions, especially about the future". Therefore, in the case of taxes, we should discount the cost of a dollar paid in the future versus a dollar paid today.

 

[Katsmeow]

If you have time, could you make a case for this viewpoint?

I think it is impossible, but I'm all eyes for your ideas.

Ha

I am not the person you were responding to but I could make case for tax rates potentially being lower in the future.

Possibility 1 -

There is a move among some to simplify the tax code, reducing overall marginal rates, and possibly collapsing them to fewer ranges. Doing this and still raising similar revenue would undoubtedly require a number of deductions to go away and/or require a hard cap on itemized deductions. While changing the tax code in this measure might be designed to raise the same revenue (or more or less), some people would end up paying more and some would end up paying less. However, it is certainly possible that the tax rates for a particular income could be less.

Possibility 2 -

I tend to think Possibility 1 is more likely. However, various people have floated the idea of some sort of VAT or national sales tax with a corresponding reduction in income tax rates. This might even be combined with Possibility 1. We all seem to think of future tax rates with the idea that US government basically collects income taxes. However, it could be that income taxes would be reduced with some sort of other tax coming in to replace some or all of the revenue.

 

[sengsational]

This might be a topic for another thread, but the OP and several follow-on posts indicate that there are some sophisticated models running! I am nowhere near where I would need to be knowledge-wise to code up an accurate model, so I just figure i-orp is ok until I can study-up. Which brings me to the question...for those of you that have a model, how different is i-orp? I know the inputs are pretty basic, but if you shoe-horn your situation, how close does it come to your detailed plan? Or is it out in left field?