Roth IRA doesn’t seem good

[MrLoco]

You don't have to pull everything out of an inherited IRA over 5 years. There is a calculator to figure out how much you have to pull out each year.

Ummmm....Yes that is what I said. I was giving an example IF Congress did away with the "stretch IRA" provision for inherited IRA's. In Sept. 2016, the Senate Committee of Finance voted 26-0 to do away with the stretch option which would have left only the 5 year rule OR the lump sum option. However, thankfully, this change never became law.

 

[Niuatoputapu]

Excellent discussion.

I agree "As pb4uski said, all things being equal its exactly the same."

I, for one, plan to continue Roth contributions and conversions, from now until 2027 (when I turn 70 and start SS and RMDs) even though I am firmly in the 22% marginal bracket. I will do so because of the following possibilities (probabilities?) in the next 30 years:

1) Federal income tax rates will increase
2) There will be SS means testing based on income (like Medicare testing)
3) I or my spouse will die leaving the survivor to pay significantly higher single status income taxes

I am happy with our tIRA balances decreasing while our Roth IRA balances increase.

 

[Huston55]

Thanks. This topic is very important to me as I'm evaluating possible conversions into the 22% bracket where I don't expect much, if any, benefit from tax rate differentials.

Conceptually, I get it. I've seen people describe Roth conversions as a kind of "transfer" from the taxable account into the Roth. In your example, the $10K tIRA is really only worth $7.5K after tax. So by moving all $10K into the Roth and paying tax from the taxable account, in effect, you have transferred $2.5K from your taxable account into the Roth, where it will live a happy, tax-free life from that point on.

The benefit is equal to the tax that would have eventually been paid on the taxable account were it not "transferred" to the Roth. I set up a simple spreadsheet that also shows the benefit is quite small for conversions up to the top of the 12% bracket and for short periods between conversion and RMD, but then increases to a meaningful amount as you convert into higher brackets with longer periods.

The more difficult problem is calculating the real-world tax drag from the taxable account. If I continue managing income and limiting conversions to the top of the 12% bracket over the next 13-14 years, then I owe nothing on QDs and LTCGs until 70 when RMDs and SS start. But my plan is to nearly exhaust the taxable account by then (the result of deferring SS and tax-deferred withdrawals). So I'm struggling to find the benefit if I never actually pay tax on the taxable account.

What's more, if I convert into the 22% bracket, I immediately pay the 27% incremental rate on QDs and LTCGs. So the net cost on conversion is higher than 22%. And I potentially deplete the taxable account even earlier.

Precisely!

We are at the beginning of the 10-12 yrs of low income before 70.5 RMDs, living from small pensions plus taxable account W/D with non-taxed LTCGs. I doubt we will use all of our taxable accounts before RMDs kick in. But, I know for sure that we’re saving 15% (actually more on anything over that bracket) tax on LTCGs by filling up the 15/12% tax bracket with LTCGs instead of doing Roth conversions.

So, it seems to me that in our situation (which I’d expect is very common), future post-70.5 taxes would have to be at least 15% higher than current taxes for us to even consider Roth conversions instead of ZERO LTCGs for the next 10 yrs. Just for comparison, let’s say the ‘blended’ rate is 18%; then, our post 70.5 tax rate has to be 33% or higher for Roth conversions to make sense...IF tax rates remain the same. And, the LTCG savings is guaranteed, while we don’t know what taxes will be in 10 yrs.

What am I missing here?

 

[pb4uski]

....The benefit is equal to the tax that would have eventually been paid on the taxable account were it not "transferred" to the Roth. I set up a simple spreadsheet that also shows the benefit is quite small for conversions up to the top of the 12% bracket and for short periods between conversion and RMD, but then increases to a meaningful amount as you convert into higher brackets with longer periods.

The more difficult problem is calculating the real-world tax drag from the taxable account. If I continue managing income and limiting conversions to the top of the 12% bracket over the next 13-14 years, then I owe nothing on QDs and LTCGs until 70 when RMDs and SS start. But my plan is to nearly exhaust the taxable account by then (the result of deferring SS and tax-deferred withdrawals). So I'm struggling to find the benefit if I never actually pay tax on the taxable account.

What's more, if I convert into the 22% bracket, I immediately pay the 27% incremental rate on QDs and LTCGs. So the net cost on conversion is higher than 22%. And I potentially deplete the taxable account even earlier.

No! The benefit is the difference between the tax currently paid on the conversion and the tax that would be paid if taken later once SS and RMDs start.

In my case, I currently pay about 8% on my conversion... some is sheltered by deductions and exemptions, some is at 10% and the remainder is at 15%.... there are also some qualified dividends and LTCG on top of that ordinary income that is 0% since I convert to the top of the 15% tax bracket.

So in 2017 I converted ~$30k and saved $4-5k [$30k*(22% or 25%-8%)].

If I didn't do any conversions, our taxes would be $0... which would be a good "feel good" but a poor decision in the long run since when the money that comes out at 8% today comes out later it will be taxes at 22%/25% or more.

 

[Cobra9777]

No! The benefit is the difference between the tax currently paid on the conversion and the tax that would be paid if taken later once SS and RMDs start...

Perhaps you should re-read the context of the discussion. We were discussing an additional benefit unrelated to tax rate differences.

 

[pb4uski]

Ok, I see now... but IMO that is a fairly inconsequential second order impact.is
For example, let's say that the taxable account earns 6% and that the tax is 20%. $2.5k accreted for 10 years at an after-tax rate of 4.8% is $3,995, the same $2.5k at 6% is $4,477 so the benefit is $482 in 10 years.

 

[Cobra9777]

Ok, I see now... but IMO that is a fairly inconsequential second order impact.is
For example, let's say that the taxable account earns 6% and that the tax is 20%. $2.5k accreted for 10 years at an after-tax rate of 4.8% is $3,995, the same $2.5k at 6% is $4,477 so the benefit is $482 in 10 years.

You just calculated one portion of the benefit which occurs PRIOR to 70. Now let's say you need some cash. The $4,477 is tax free. The $3,995 is only $3,196 after tax. The net spendable difference at 70 is $1,281 on the original $2,500. For a 15 year time-frame, it goes to $1,951. The benefit also gets bigger as you convert into higher brackets because you are "transferring" larger amounts from taxable to Roth.

Also, it may be a second order impact in your situation. But the discussion was about equal rates at conversion and RMD. So in that context, it is the only benefit. I'm evaluating conversions into the 22% bracket where I expect little or no rate-related benefit.

All that said, I'm still struggling with the concept because the real-world tax consequences in my taxable account are not like the simple examples. I had assumed no benefit prior to 70 (0% on QDs and LTCGs). And at 70, I expect the taxable account to be nearly gone. As Animorph suggested in post #13, I think this could be a significant impact if you are already paying tax on a very large taxable account, and you're converting into the 22% and 24% brackets, and you have a long time period from conversion to RMD.

 

[Plantman]

I still say the Roth is superior because of the flexibility. If I would like to withdraw a million dollars at once from my account it would certainly be beneficial to have it in a Roth instead of a traditional IRA!

 

[pb4uski]

You just calculated one portion of the benefit which occurs PRIOR to 70. Now let's say you need some cash. The $4,477 is tax free. The $3,995 is only $3,196 after tax. The net spendable difference at 70 is $1,281 on the original $2,500. For a 15 year time-frame, it goes to $1,951. The benefit also gets bigger as you convert into higher brackets because you are "transferring" larger amounts from taxable to Roth.

Also, it may be a second order impact in your situation. But the discussion was about equal rates at conversion and RMD. So in that context, it is the only benefit. I'm evaluating conversions into the 22% bracket where I expect little or no rate-related benefit.

All that said, I'm still struggling with the concept because the real-world tax consequences in my taxable account are not like the simple examples. I had assumed no benefit prior to 70 (0% on QDs and LTCGs). And at 70, I expect the taxable account to be nearly gone. As Animorph suggested in post #13, I think this could be a significant impact if you are already paying tax on a very large taxable account, and you're converting into the 22% and 24% brackets, and you have a long time period from conversion to RMD.

Your first part isn't right... the $3,995 is after-tax and is totally available since it is in a taxable account and the taxes in taxable account income have already been paid... the other part of $4,477 was being used only to calculate the second order impact of taxes in the taxable account.
========================================================
The example you used was $10k of conversion with 25% tax rate or $2.5k in tax.

Let's say that you start out with $2.5k in taxable funds and $10k in tax-deferred.

Scenario 1: you convert, use the taxable fund to pay the $2.5k in taxes and end up with $0 taxable and $10k in the Roth. After 10 years at 6%, the Roth is worth $17.9k

Scenario 2: you don't convert. The $10k of tax deferred grows at 6% to $17.9k. The $2.5k of taxable grows at 4.5% after tax to $3.9k. Your withdraw the $17.9k and pay $4.5 in tax and end the day with $17.3k ($17.9k - $4.5k + $3.9k).

The difference is $0.6k. That difference is also the difference between $2.5k grown at 6% (no-tax) of $4.5 and $2.5k grown at 4.5% (after-tax) of $3.9k and is the value of avoiding taxes on the taxable account for 10 years because it was invested in the Roth.

Essentially, even if the tax rate is the same the Roth wins because it avoids taxes on income earned on money in the taxable account... by using taxable funds to ay the taxes it is like reinvesting the money in the Roth and that money no longer is subject to taxes.

Note: in this last example I change the tax rate on the taxbale account to 25% to be consistent with the 25% used in the prior example.

 

[kaneohe]

...........................
The difference is $0.6k. That difference is also the difference between $2.5k grown at 6% (no-tax) of $4.5 and $2.5k grown at 4.5% (after-tax) of $3.9k and is the value of avoiding taxes on the taxable account for 10 years because it was invested in the Roth.

................................................

nice intuitive insight above.........................I wonder if the difference might be somewhat smaller.......third order effect. If growth rate is 6%, 2% might be dividends taxed currently and 4% might be unrealized CG. The 2% taxed at 15% is 0.3% so wouldn't the compounding be at 5.7%. Of course, then you have to deal w/ the CG at the end and the basis increase from reinvestment.
I think you brought that up some time ago.

 

[pb4uski]

Agreed... I was just trying to keep the example consistent with the 25% tax rate that Cobra used in the initial example for consistency.... but I agree in reality the benefit would be the difference between the actual expected tax rate on the taxable money used to pay the tax on the Roth conversion and the 0% once it is in the Roth.

 

[Brat]

One advantage few have mentioned is that when you are in the RMD phase of life your tIRA balance is lower while the converted money in the Roth grows. Investment wise a conversion may be a push when taxes are figured in. That isn't a big deal for the first 10 years but once you are 81+ your RMDs ramp up significantly.

 

[Cobra9777]

Your first part isn't right... the $3,995 is after-tax and is totally available since it is in a taxable account and the taxes in taxable account income have already been paid...

You're right. I was overstating the tax due at liquidation.

nice intuitive insight above.........................I wonder if the difference might be somewhat smaller.......third order effect. If growth rate is 6%, 2% might be dividends taxed currently and 4% might be unrealized CG. The 2% taxed at 15% is 0.3% so wouldn't the compounding be at 5.7%. Of course, then you have to deal w/ the CG at the end and the basis increase from reinvestment.
I think you brought that up some time ago.

I agree. I just set up a table starting with $2.5K in a taxable account, growing at 6% per annum, of which 2% is dividends taxed currently at 15%, reinvested after paying tax, and the total growth (CG) taxed at 15% at the end of 10 years. Total tax paid was $375 (dividends along the way plus CG at the end). I think this is perhaps a more realistic measurement of a "typical" benefit from each $2.5K invested in the Roth rather than taxable.

Pretty small, but it is for *each* $2.5K of tax paid on Roth conversions from the taxable account. For someone looking to incrementally fill up the 22% bracket with Roth conversions, that's an additional $87.6K conversion at 22% or $19K tax. Doing that for 10 or 15 years would definitely shift a lot of taxable funds into Roth, although the benefit gets smaller each year as the period between conversion and RMD decreases.

In the past, I never considered converting above the top of the 15% bracket, even though that would still leave a very large tax-deferred balance at 70. But with the new 22% rate, and the possibility of reverting to 25% just in time for our RMDs, I think it might make more sense now. This additional benefit, while small, makes the whole decision a little more palatable. Definitely worth some additional analysis.

 

[pb4uski]

.... In the past, I never considered converting above the top of the 15% bracket, even though that would still leave a very large tax-deferred balance at 70. But with the new 22% rate, and the possibility of reverting to 25% just in time for our RMDs, I think it might make more sense now. This additional benefit, while small, makes the whole decision a little more palatable. Definitely worth some additional analysis.

+1 here but I'll likely wait a few years until we are 65 and on Medicare and redomesticate to Florida as state income taxes would be an additional drag.

 

[CardsFan]

The comments, calculations and analysis in this thread seem to confirm my current plan:

1. If I do nothing, RMD's will definitely be in at least the new 22% bracket.

2. Convert to the top of the new 12% bracket. A no-brainer, since virtually all contributions avoided 25-28% tax rate

3. Conversions in the 22-24% bracket COULD help, but we are talking low single digit benefits, at best. So, for now I will take a pass, UNLESS:

4. If we see a significant pull back in the market (20-30%) then converting even in the higher brackets could produce a big benefit, assuming the markets eventually return to current levels. Could call this market timing, I guess.

The only other big advantage of converting in the higher brackets is to provide a tax avoidance strategy for DS. As much as I love him, I won't feel too bad if our legacy pushes him into the 24% tax bracket with a stretch IRA (or maybe even higher if this is eliminated). That just means he won the game too (and he didn't even know he was playing)

 

[AginMS]

I appreciate the discussion and pb4uski's analysis throughout this thread as it helps drive home the point that, all things being equal, a ROTH and an IRA will have the same outcome.

It doesn't matter if the marginal tax rate is the same.

Let's take your first year contribution of $5,000 and let's say the tax rate is 20% to make the figuring easier. You have earned $6,250 and have two choices: defer $6,250 and pay no tax now or pay $1,250 in tax now and put $5,000 in a Roth.

Both accounts earn 7%/annum.

30 years later, the tax-deferred account is worth $47,577 [$6,250 * (1+7%)^30] and you take it out and walk away with $38,061 after paying $9,515 (20%) in taxes.

30 years later the Roth is worth $38,061 [$5,000 * (1+7%)^30]

Mathematically, unless your tax rate is different then it doesn't matter. Now if your tax rate in retirement is 15% then the tIRA is worth $40,440 so you are better off having done the tIRA than the Roth. Conversely, if you have been spectacularly successful so your tax rate in retirement is 25%, then the tIRA is only worth $35,682 and you would have been better off with the Roth.

However, in the example above, the tIRA got the entire tax deferred amount deposited ($6,250), but the current limit is $5,500 so the additional $750 would need to have been invested elsewhere which would change the outcome.

But this isn't really the point I wanted to make about ROTH vs. tIRA. For many young investors (i.e. my 18 yo DD) I think it is a HUGE win just to get them starting to invest in the first place. But I would bet most young adults starting out just have a number in mind that they would like to invest, say $300/month, and just because the money is tax deferred or not isn't going to change the amount they have in mind because there is some emotion (vice hard logic) in a number that "feels" right for investing.

If they can be convinced to max their ROTH vice tIRA limit of $5,500/yr, then I would bet the vast majority would come out ahead because 1) their marginal tax rates are likely to be lower as a young investor than as a retiree (I know, I know, no guarantees), and 2) most young investors (heck, most people) won't invest the tax deferred portion of the tIRA. Instead, that tax deferred income is simply absorbed into their budgets and spending plans. So while there is some short term value to be gained in the tIRA (i.e. the consumable "stuff" you buy vice having the ROTH contribution taxed), generally speaking I think the ROTH makes a lot of sense for the young and average investor.