Roth Conversions - Much Ado About Little?

[sengsational]

Like comic acting, in retirement planning timing is everything.

James Welsh wrote another paper recently. You can get it by going to i-orp.com, scrolling down to the bottom, and click on "A new white paper about IRA to Roth IRA conversions is posted on the ORP". You might consider dropping a fiver in the tip jar while you're there.

The paper compares scenarios with and without Roth conversions, and spells out, for those examples, what the "levers" are. In other words, what makes Roth conversions more worthwhile, and what makes them less worthwhile. Of course you could feed your own information into the i-orp page and run one with and one without the Roth conversion checkbox, and see how your situation measures up. As an aside, that's worth five bucks, for sure, probably worth more like $50, actually. I have no axe to grind except I think this tool is awesome and don't want a smallish server bill to cause the tool to evaporate. From my first hand experience, I can say that just about nobody clicks the donate button for free sites. Maybe I'm wrong, since this site has a lot of "good tippers" when it comes to restaurants, maybe that translates to PayPal donate buttons too. James deserves at least as much as your waitress, n'est-ce pas?

Anyway, back to the content of the paper...

Our results are that partial conversions increase disposable income by less than 1% in most situations...the financial benefits of doing partial IRA to Roth IRA conversions may not warrant the extra paperwork.

Of course, if you thrive on optimization, and you take that away, life would become so much less interesting, lol! This is where I fall on the question of whether to bother with conversions

One thing that the paper called to my attention was the break-even age for the two scenarios (the Roth conversion approach and the non-conversion approach). For the example, about 2/3 of the time between 65 and 92 was spent in territory that, should you die, you would have been better-off not doing the conversions indicated by the optimization.

It should be no surprise if you've read the i-orp help and/or James' comments on the subject, the conclusion of the paper:

The conversion decision should probably be made based on considerations other than financial gain.

 

[jabbahop]

thanks for the reminder to donate. i too value the tool so sent along a few bucks.

 

[robertf57]

Thanks for the heads up on the discussion. I will have to look it over. Of course, the i-orp calculator doesn't model significant increases in tax rates which may be imposed in the future or the financial benefit to heirs of a roth inheritance.

 

[unclemick]

Katrina and no house/camp insurance because part of the structure was over water. Also did a conversion when a tornado took the roof off and not all losses where covered.

And yes I love ORP and tipped them $100 a while back cause I'm cheap.

heh heh heh - you reminded I should 'tip' again cause it's one of my favorites.

 

[Gotadimple]

Thanks for the heads up on the discussion. I will have to look it over. Of course, the i-orp calculator doesn't model significant increases in tax rates which may be imposed in the future or the financial benefit to heirs of a roth inheritance.

If James could model the future, all of us would have been following our I-ORP results to the last digit. He can't and advises you in the notes to the reports to use the results as a guide or pattern for action.

- Rita

 

[GrayHare]

If income taxation remains much the same as is during the coming decades, Roth is a wash for most people. It's the less common financial situations that are the interesting Roth applications.

 

[BBQ-Nut]

I have constructed my own modeling tool in Excel that lets me play the 'what if' game for several factors, including Roth conversions post-FIRE.

It uses current tax tables/rates and I can change rates of return, amounts to convert, accounts for RMDs, SS, etc.

What I have found for the DW and myself accounting for our different assets (after tax, deferred and tax free) is that when combined with our pensions and SS, our accounts are of sufficient size that doing Roths only ends up protecting assets at 'end game' and does not result in any significant changes in overall portfolio balances.

Of course, the big gamble is that taxes remain the same.

All bets are off if there are any major changes/increases to the tax structure or rates.

Assuming Roths would still be considered 'tax free' many many years from now, then I suppose one justification for doing partial conversions would be to 'future proof' asset values in the event that tax rates go up significantly.

 

[RunningBum]

Our results are that partial conversions increase disposable income by less than 1% in most situations...the financial benefits of doing partial IRA to Roth IRA conversions may not warrant the extra paperwork.

Yet the conclusion states:

Table 1 indicates that IRA to Roth IRA conversions increase disposable income from 0.6% to 3%.

I browsed through the paper and don't know if those apparent differences are really at odds or not. It's pretty complex. I'll probably stick to my own calcs for my own situation. It's been awhile since I've run those so I can't say if my benefit is different.

I'd also note that the paper talks about a 65 yr old retiree starting partial conversions at that point. Many of us are stretching our conversions over a much longer time period. I would guess that helps.

 

[Midpack]

Good read, thanks for pointing it out. Direct link http://www.i-orp.com/ModelDescription/Conversions.pdf

Conclusions

Deciding to do partial IRA to Roth conversions is not the obvious choice that it intuitively may seem to be.

Table 1 indicates that IRA to Roth IRA conversions increase disposable income from 0.6% to 3%.

Figure 3 shows that conversions protect against future tax increases by reducing the taxable balance in the IRA later in retirement. The open issue is how much of a tax increase and to which tax brackets, warrant conversions.

Table 3 shows that there is a small penalty to be incurred by directing savings to the Roth IRA with the intention reducing the tax burden for the heirs.

There is, as yet, no rule of thumb for identifying the optimal age (Figure 2) to lower IRA distributions to the top of the 10% bracket and increase Roth to compensate.

Given:
1. the aversion to using retirement savings to prepay taxes in a higher tax bracket early in retirement,
2. the potential reduction in the size of the estate in mid plan,
3. the withdrawal timing issue, and
4. the minor financial benefit from doing partial IRA to Roth IRA conversions,

The conversion decision should probably be made based on considerations other than financial gain.

 

[Koolau]

Assuming the taxes are a wash (big assumption - since we don't know what future tax rates will be) there are still some advantages to Roths. As alluded to, the passing to future generations can be more favorable. There are no RMDs to be "forgotten" by us "old" people. The biggest (potential) advantage is increasing the "value" of tax "free" earnings through conversion. The cost to do so is to use already taxed money to pay the taxes due upon conversion. One could make a "wash" argument with this as well, but if you choose investments well, you could reap significant benefits in the long run. It's theoretical until you pull it off so YMMV.

 

[Independent]

When I tried I-orp, I input enough after-tax assets to cover my spending for the first three years.

I-orp used the after-tax money first, then shifted to the tax-deferred. But, the "tax" column in the output didn't change materially. I should have seen a large increase.

Is it using some very simple tax rule?

 

[clifp]

Assuming the taxes are a wash (big assumption - since we don't know what future tax rates will be) there are still some advantages to Roths. As alluded to, the passing to future generations can be more favorable. There are no RMDs to be "forgotten" by us "old" people. The biggest (potential) advantage is increasing the "value" of tax "free" earnings through conversion. The cost to do so is to use already taxed money to pay the taxes due upon conversion. One could make a "wash" argument with this as well, but if you choose investments well, you could reap significant benefits in the long run. It's theoretical until you pull it off so YMMV.

I agree. As long as withdrawal from Roth's are not treated as income, there will be other advantages to having a good size Roth. ACA subsidies, SS taxation, and probably in the future various means testing on benefits such as Medicare or SS. Now I think there is a reasonable probability a future will treat those as loophole and try and plug them.

That said I've been pretty adverse to paying taxes now for a benefit in the future. I've also been almost burned by Roth conversions in the past, so I have NOT been doing conversions consistently. So it is nice to see that I'm probably not giving up much.

 

[hnzw_rui]

I browsed through the paper and don't know if those apparent differences are really at odds or not. It's pretty complex. I'll probably stick to my own calcs for my own situation. It's been awhile since I've run those so I can't say if my benefit is different.

True.

I'm planning on using Roth conversions to smooth out withdrawals from tax deferred accounts. It would be good to have "tax-free" funds in case of unusually high, irregular expenses (e.g. new car, extra fancy vacation, home repair, etc).

 

[RetireBy90]

Assuming the taxes are a wash (big assumption - since we don't know what future tax rates will be) there are still some advantages to Roths. As alluded to, the passing to future generations can be more favorable. There are no RMDs to be "forgotten" by us "old" people. The biggest (potential) advantage is increasing the "value" of tax "free" earnings through conversion. The cost to do so is to use already taxed money to pay the taxes due upon conversion. One could make a "wash" argument with this as well, but if you choose investments well, you could reap significant benefits in the long run. It's theoretical until you pull it off so YMMV.

Agree with the quote - thanks to all that have commented. However, for me a large benefit to a ROTH is the elimination of unknown future taxes. I model what my future income and taxes will be, but there are many unknowns. I won't know how much my RMD will be each year because I don't know what the value of my trad IRA will be in 10 years. If you take a risk, you should be compensated for it, and if I leave $$ in my traditional account I take a future risk, don't I ?

So for me the benefit I care most about is that I buy flexibility. Insurance if you will against increase in taxes. Of course, taxes could go down or worst would be after pre-paying all these taxes by converting, that Feds go to a VAT or national sales tax and drop the income tax. Time to pay again

 

[hnzw_rui]

So for me the benefit I care most about is that I buy flexibility. Insurance if you will against increase in taxes. Of course, taxes could go down or worst would be after pre-paying all these taxes by converting, that Feds go to a VAT or national sales tax and drop the income tax. Time to pay again

Which is a great case for not going all in. As with asset allocation, diversification is key.

 

[RetireBy90]

Which is a great case for not going all in. As with asset allocation, diversification is key.

I hope to have it 50/50 by 70, and to live off RMD plus pensions and SS, leaving Roth to the kids.

 

[Rothman]

If you are retiring early and living off taxable accounts until SS or pension kick in why wouldn't you do Roth conversions at tax rates below where you will be withdrawing later? Now if you're managing to ACA subsidy limit then that's your cap but why leave money unconverted, do you think tax rates could be lower in the future? I expect we'll find many people whose portfolios will do better than they expected getting RMDs beyond their spending needs and growing their taxable account. Now I am not arguing to extreme, you'll still leave a large tax deferred balance too, account diversity provides more options for your withdraw needs.


Sent from my iPad using Early Retirement Forum

 

[RetireBy90]

If you retire early, why not stretch out the traditional withdrawals? First, any withdrawals needed to fund retirement; second any conversions to Roth; third withdrawals to keep tax rate lower than if you wait till RMD time.

 

[Sunset]

Agree with the quote - thanks to all that have commented. However, for me a large benefit to a ROTH is the elimination of unknown future taxes. I model what my future income and taxes will be, but there are many unknowns. I won't know how much my RMD will be each year because I don't know what the value of my trad IRA will be in 10 years. If you take a risk, you should be compensated for it, and if I leave $$ in my traditional account I take a future risk, don't I ?

So for me the benefit I care most about is that I buy flexibility. Insurance if you will against increase in taxes. Of course, taxes could go down or worst would be after pre-paying all these taxes by converting, that Feds go to a VAT or national sales tax and drop the income tax. Time to pay again

Or, you do all the conversions over a few years paying the taxes and then the stock market drops 50% meaning you could have converted everything at 1/2 the tax.

 

[BBQ-Nut]

Money has to come from somewhere to pay the tax on the amount of the conversion.


So, depleting after tax assets or other accounts may be at risk, if needed to cover expenses, more so during a bad sequence of returns that it may never recover from.


There may be a limit to how much one can convert because of the present taxes due.


It all needs to go into the 'equation' to determine what is right for your situation.


I don't think there is any one 'right' answer.

 

[RetireBy90]

Money has to come from somewhere to pay the tax on the amount of the conversion.


So, depleting after tax assets or other accounts may be at risk, if needed to cover expenses, more so during a bad sequence of returns that it may never recover from.

I don't think there is any one 'right' answer.

While I agree tax bill has to be paid from somewhere and there isn't any one 'right' answer, no matter when you take the money from taxable IRA/401K you will have to pay the tax.

2 points on this, and it is entirely personal to me, not the right answer for others.
1) If I take distributions from tax deferred account say from 60 to 70 then I MAY be able to pay a lower effective rate overall considering I will be required to take RMD at age 70. I will have a rental paid off when I turn 70 so I'll be paying a higher effective tax rate, and won't need the RMD money.
2) As others have stated, diversify between tax deferred and ROTH and taxable gives you options.

 

[NoiseBoy]

Thanks for posting the link. I read the paper and the charts and graphs really helped me understand the dynamics of paying taxes up front in order to differ or even avoid taxes later. Some food for thought in the article for me, as well as a reminder to run I-Orp. One issue I have with the calculator is that I have 2 pensions, one that I plan to start at 55, and one that I can't start until 62. Any ideas on how to model a different pension start age in I-Orp?

 

[sengsational]

Is it using some very simple tax rule?

Not real nuanced, as far as I can tell. There is a table in the output that shows how much was untaxed, how much in the 10% bracket, etc. You might be able to use that to see if it's working as expected. My one wish is that we could enter the PPACA PTC amount, which, for early retirees, makes a pretty big difference in "taxes" (they pay ME!)

... large benefit to a ROTH is the elimination of unknown future taxes.

Good point. The assumption of tax increases through time might be a nice addition to the model.

Feds go to a VAT or national sales tax and drop the income tax. Time to pay again

I hope there's an exclusion if they go that route. I think there would be riots by the AARP crowd if there wasn't a rebate for taxes already paid.

I don't think there is any one 'right' answer.

That's basically what the paper says. But if you look across a lot of scenarios, and accept the model's assumptions (admittedly kind of a big "if" when it comes to taxes staying the same), many of them don't have much payback.

Any ideas on how to model a different pension start age in I-Orp?

You could calculate a present value of both pensions, then calculate an annuity starting in the year of the later one....that should be the conservative way to get that into the model.

 

[hnzw_rui]

I hope to have it 50/50 by 70, and to live off RMD plus pensions and SS, leaving Roth to the kids.

Sounds like a decent plan, then.

 

[samclem]

It seems to me that future investment performance is >the< biggest unknowable in this "how much should I convert?" situation. It impacts the size of any future RMDs and the taxes to be paid on them. But, I know a few things:
1) If my investments perform poorly compared to my present expectations, I'll need the dollars to meet spending requirements and my marginal tax rate will be the same as my present rate or lower. I will be sorry if I converted too much and gave the dollars to the IRS in my earlier years.
2) If my investments have higher-than-expected returns, then I'll pay extra taxes on funds I didn't convert earlier, but the high returns mean I'll still have more money to spend than I'd planned on, so paying the taxes at this later time is not very painful. And--the uncertainty about portfolio returns and required spending decreases each year (as we approach our appointment with the grim reaper. A 60 year old needs to worry about returns/expenses for long timeline, a 90 year old--less so).

So, given the relative marginal utility of the dollars in each scenario, I bias toward converting less rather than converting more. The expected after-tax values in each case are not of primary importance, it is the expected utility of the remaining funds and the different "pain" of the tax bite. In my case, I will probably convert to the top of the 15% bracket (our present bracket), but not beyond that.